 Welcome everyone. I'm Amal Andralis. I'm the Dean of the School and I'm really thrilled to be having you all today in our school and also hosting this amazing conference. You should know that it's, well, we don't sell but if we did sell it would be sold out. And just the response has been incredible and it's so exciting to have scholars and practitioners from across the disciplines of technology and the built environment and I see alumni and, you know, thinking about change in fact and how we deal with change. Everything is moving so fast as systems that we once established are being transformed and it's not clear whether it's always for the better. And so for us to come together to think about these changes and their impacts and how we can kind of redirect them and harness them and make them visible and understand them and present them in a new way and find new modes of collaboration I think is what we're all kind of here today to assemble around urgent questions and in general I think this is the urgency that we find ourselves in where institutions and disciplines are both unable to adapt and yet this is what we have to transform and retool. And I'm particularly excited that the conference was really masterminded by Leah Meisterlin who is, as you know, assistant professor here at the school but even more importantly I think Leah is one of the kind of seed bridges that we have, you know, sort of put forth in the hope of stitching back together architecture and urban planning and this body that was once divided needs to act as one again in new ways and I think Leah's, you know, unbelievable contribution already even within a few years to single-handedly transforming the kind of thinking and work of both architecture and planning students makes me hopeful. So please welcome Leah Meisterlin. Good morning. I'm so excited. Okay, I'll admit I've been excited for today in many ways much longer than it takes to organize a conference. I want to start by thanking the school and the dean Amal especially for the support you've given to this conference and that support is very much appreciated and very deeply felt. I want to thank Laila Kattelier and her entire events and communications teams. None of this is possible without your enthusiasm and your wizardry and I want to thank all of our participants from whom you will hear today, participants, contributors and of course the moderators as well for your time, your attention, your interest, your work and your care because honestly, well rather I'd like to start as honestly as I can by saying that I don't exactly know where to begin. I do know that somewhere at some point in the last decade I lost count of the many irrevocably consequential changes in the digital material landscape of cities and I lose track consistently trying to follow the sociopolitical cultural and economic processes that flow across through and against those landscapes. I have lost count of the devices and the disruptions, the platforms and the predictions. I have lost track of the algorithmic and artificial, automated and autonomous. I suspect some of you are like me in this regard. I am losing count of the new ways to count and by that mechanism the ways we determine who counts. I'm afraid I'm losing track of everything that's designed to track and of course we're here because in both of those cases we're really talking about people. We're talking about technologies designed to count and track people. We're here to talk about digital technologies and cities which might mean very many different things today but for me at least right now I mean the collection of connective platforms, computational processes and informational infrastructures that are reorganizing, restructuring and in many cases simply accelerating social and capital relations within cities and across regions and specifically accomplishing this by changing the means by which places and people, neighborhoods and communities are designed, planned, conceived, imagined, operated and represented. So here's a quick, really close to home and admittedly advantageously current example. Think of we work six months ago and think of we work today. Whatever happens internally for that organization in New York we work is not exactly a start-up but a technological intervention into the very code of the city's planning and its socioeconomic reproduction. As New York City's largest tenant it is worth asking what happens if we work doesn't make its rent and whether New Yorkers will have to decide that this particular firm is now too big to fail. I can reframe that question as one of technological and built scale. It's worth asking whether commercial rent arbitrage predicated on precarious labor and facilitated by fastidious data collection and analysis that is counting and tracking, right, has scaled beyond the units of desks per week and conference rooms per hour into something not only larger but qualitatively different, something that involves and implicates all of us. Put yet another way, where's the line? What is the scalar threshold between space as a service and space as a responsibility? We work as just one example. Think ride-sharing and autonomous vehicles, consider e-commerce logistics and personal GPS enabled wayfinding and routing, crime mapping and predictive policing or social media-based targeted advertising. Their internal logics, their scale and reach, their math, their hardware and their impacts on human urban environments are not dissimilar. Among their commonalities, they all involve spatial technologies of sensing and extraction, extracting information as commodity, as material, extracting attention, labor and time, and particularly in a spatially regulated democracy, extracting votes. We know that now. Okay, so before this gets too dark, I don't think we're here to talk about the good, the bad and the optimized. I won't speak for everyone, but I'm not here personally to vilify smart city sales pitches, nor am I here necessarily to accept techno-positivism or the techno-optimism that usually follows it. Rather, I am so excited, which means I will stop talking soon so that we can get into it. I'm so excited because across today we will hear about digitality, high-powered counting in terms that involve and implicate all of us in research, in design, in action and in outcome, in the messy, inefficient, ambiguous, conflicted, diverse terms of urbanism, plural. The pluralism of cities, the co-location and simultaneity of different communities and their spatial practices and priorities presents challenges that plainly digital technologies cannot resolve by definition. And thus, it recalls questions and challenges that are absolutely not new. Those of access and distribution, inclusion and marginalization, equity and justice, these are questions of responsibility and obligation. So, plan for the morning. I'm going to very quickly introduce the moderators for the two morning panels, who will in turn introduce you to our incredible and generous speakers. If you haven't already, please grab a program, which does include full bios and abstracts. They're up there. We will have two panels, like I said, before we break for lunch, and each of them with three presentations followed by a discussion. I'm going to go in reverse order. Our second panel on systems of representation will be moderated by Mark Wasuda, co-director of the Critical Curatorial and Conceptual Practices in Architecture Program, here at GISA. Among his many other super-interesting projects is ongoing research and exhibition work, control syntax, examining smart city forms and formulations, and the processes of computational urbanism they prescribe and otherwise engender. Before that panel, I would like to invite up Malo Hudson, associate professor in urban planning, also here at Columbia GISA, and director of the school's Urban Community and Health Equity Lab. His work sits at the intersection of community development and health equity, with a specific mind toward racial and ethnic inequalities, and while not in your official bio, I also think your Malo's recent work on circular economies were warrants mentioning, because it could certainly be of interest to our discussion. So thank you very much. Thank you for the kind introduction and for inviting me here. As Lea said, I'm Malo Hudson. I want to welcome you all to Columbia. And I wanted to start off by first thanking Dean Amal Andrews. I think that when you think about urban planning and architecture and urban design and thinking about the built environment, overall you have to think about innovation and where we're going as a society. And so one of the things I'm so proud about being a part of GISAAP is the vision in this department, or in the school I should say, and how we think about innovation, technology, and so forth. And so I look forward to today's discussion. I also want to thank my colleague Lea Meiserlin, who's been very humble, but I would have to say it's a pleasure to watch you as a young scholar and put something so important like this together, and having the leadership to do so. And I know it takes a lot of time, but I hope today you can see all of that hard work has paid off. So anyway, again, I want to thank you for your time. So as I was thinking about moderating this panel, I thought about my own work and what not. And 20 years ago, I started my doctoral program at MIT. I had the opportunity to work with people like William Mitchell, Stephen Graham, and Manuel Castells. And many of us as young doctoral students were thinking about the growth of technology, certainly Silicon Valley was taking off. And we didn't talk about digital urbanisms then. We were thinking about those things. We didn't have it quite defined like it is today, but we're thinking about information technology. And what role would technology play in altering our daily lives in terms of where we live, where we work, thinking about mobility in general, thinking about inequality, opportunity. There was a big discussion about the digital divide and all of those things. And so I went back last night and looked at our old journals, our information technology in place making. And so I'd have to say we've come a long way. And so the panel that I'm going to moderate is on infrastructures and digital materiality. And part of that again goes back to many of the same themes we were talking about before, but certainly much more advanced. But thinking about the role of digital technologies in our everyday lives, whether that be sensors, whether that be big data, whether that be privacy, security, all the things that we think about now. And so I would like to bring up our first panelist who we'll talk about. I'll introduce one by one, but that will begin to move us in this direction. So the first panelist is Narissa Moret. And I'll give it, their bios are in the booklet, but I'll say a few things. But Narissa is a strategic planner and project manager who has worked extensively with both public and private sector clients on urban innovation, parks and open space, coastal planning and real estate, and economic development projects. At present, she's an associate director of America's most innovative urban development project, Sidewalk Toronto. And part of this with Sidewalk Toronto is really thinking about the role that technology, new technologies can play to improve the lives of urban citizens, thinking about the digital, the physical, and also policy and financing. And so the case she'll present today will really dive into issues around urban mobility and thinking about new systems, sustainability and stormwater infrastructure, and outcome-based building code. So with that said, please give a warm welcome to Narissa Moret. I'm going to go a little shallow and wide today rather than deep, so hopefully that will be okay. But I think that the project that I'm working on in Toronto can introduce so many of the ideas that are going to be talked about not only in our panel, but later on this afternoon. So I wanted to kind of give you a touchstone across the whole of the project. But also I wanted to talk a little bit at the beginning just about Sidewalk Labs, where I work what we look for and think about in projects as well as then introduce the Sidewalk Toronto project as a case study. So it's no surprise to the crowd here today that cities are ever-facing more struggles and challenges in quality of life as we face continued urbanization around the world. And in response to this, we've always looked to technology through the ages. But we know from history that technology can have both positive and negative outcomes. And that's absolutely still true for what we're facing now in this next digital revolution. We have to think really carefully about the impacts that what we do are going to have on the communities and society around us. So Sidewalk Labs was actually formulated to think about these questions. It's a company that's often sort of characterized as a smart city company. But actually what's interesting is that the vast majority of people working at Sidewalk are geeky urban planners and architects and engineers. And they're mixed together with technologists. But these are people most of us coming from having worked for the cities with cities. And we are sort of looking for that spot, looking for opportunity without bringing on the negative consequences. So we do this by looking at all areas of innovation. So today we're talking about digital innovation but in fact the company looks at innovation across the board. So we look at materials innovation, we look at policy and financing all of the tools that you have to look at in the toolkit to think about quality of life. Because the way we think of ourselves as a firm is actually as a firm that's looking at increasing quality of life for citizens. It's not about putting some digital layer on with nothing else it's really looking at the outcomes that all of these different kind of new innovations can bring in cities. So we've organized the company in what we call five pillars and this is important because all of these parts of projects of city making we think are equally important. So it's social infrastructure, buildings and housing, mobility, public realm and sustainability with an emphasis on climate positive. We don't think that projects can or should be undertaken without all of these areas being investigated. Again it's not just about sticking some sort of digital technology on top it's about looking at how can all of these different areas be taken together to really improve outcomes for citizens. So how do we think about sort of implementing that or as a framework for projects we think of these five pillars as sort of different physical layers in the environment with the digital layer is just one more of those and all together these sort of six different elements together are a framework for looking or approaching a project. So what does that mean in terms of a real project? I'd like to talk a little bit about Sidewalk Toronto in this context. So Sidewalk Toronto is a project it's 15 acres of land on the eastern waterfront in Toronto. It's on the very far east of the waterfront but still that being said it's only 15 or 20 minutes walk from the major transportation hub in downtown Toronto. So it's very close to the downtown area. It's a brownfield landfill ex-industrial site with almost no tenancy on it at the moment. It's owned by Waterfront Toronto which is a tripartite government organization with representatives from the city province and from the federal government and they've been developing the waterfront over the last 15 or 20 years. They're quite progressive so they've been doing a lot of progressive development of the waterfront in downtown Toronto and so with this project which is one of the last parcels of land they were really looking to kind of move the bar on innovation with the development. So they went out with an RFP for the site particularly looking for an innovation and financing partner. So we responded to the RFP and we won in 2017 right at the end of the year and what did we win? We won the right to actually prepare a plan for the site. So that plan has to be ratified by government, by both Waterfront Toronto and the city as well. But that's what we've been doing is preparing that plan over the last year and a half and it was presented to Waterfront Toronto in June of this past summer. One of the reasons we are interested in the project of Waterfront Toronto is that Waterfront Toronto's goals for the project are very similar to what we're interested in. So they're interested in job creation, housing, sustainability mobility and urban innovation. It felt like a really great fit for us. They're looking holistically at the project and holistically at the outcomes that they want by developing this parcel of land. So how do we approach a project from the beginning? Well because we are a kind of geeky group of urban planners and so on we do look at it just from a normal lens of urban design and urban planning at the beginning. We look at what the relationship of the site is to the surrounding city. We look at the connectivity. We look at what the public spaces are going to be. In Keyside, the name of the site is the Keyside site. We're excited to have three new public spaces that we're going to be bringing online and so our design team and landscape architects have been looking at how to design those spaces. But today since we're talking about digital layers and the digital architecture what I wanted to do is for each of these different pillars like for public space sort of talk about the digital layer that we see for each of these. So for example for public space we're really interested in being able to enable more people more of the time to be outside and to be using public space. So we think about that through basic design but we also think about the digital layer and how that can help to realize that outcome. So whether it's something like whether mitigation structures that are tied to real-time data about the weather that can deploy different structures and protect different areas of the site to make it more usable outdoor space. Whether it's flexible multi-space areas that have a digital architecture underlayer that make the space flexible so that communities can adapt that space in many different ways whether it be today or next week or 10 years in the future when the needs of the community change. Whether it's ubiquitous connectivity and Wi-Fi to kind of lower the digital access and digital divide for people in the neighborhood. Whether it's developing tools so that the community can self-manage its own programming, advocate for what they want in spaces. These are all different ways, are different tools we look at for using a digital layer to actually provide those outcomes that we look for. So mobility is the same way. The actual most important backbone of mobility for the site to us is transit. In any dense urban area having great public transit is the most important thing we feel. Here in Toronto what we're actually pushing there is new financing vehicles for financing extension of the LRT through neighborhoods to the west of us and then through our site. So that's actually not a digital layer it's actually a financing tool but I want to mention it because it's the most important thing to us. But then we're also interested in lots of other digital infrastructure developments as well. So for us the reduction of public, sorry, the reduction of private car owner trip and private car trips is one of our most important outcomes for mobility. So how do we do that? We look to obviously support public transit but also to make active transportation more attractive. So we do that for example through adaptive traffic signals and bicycle green waves. Adaptive traffic signals can help speed the LRT faster so give it better performance times. It can make sidewalks safer towards a vision zero goals by having sort of adaptive traffic adaptive crosswalks in the sidewalk and it can use things like the bicycle green wave to speed up travel by bicycles and make their transit through signals more safe. One of the things I'm particularly interested in is our curb management, a dynamic curb. We're planning for the streets in Keyside to be curbless also to have no parking on the street and no parking lanes. So here in the bottom two pictures on the slide you can see what we traditionally thought of sidewalk. On the left there's a car, it's pulling in, it's a drop off and pick up zone that's variable used in high peak times for vehicles but then in off peak times you can use it as public space so you're reclaiming space back as one of the goals is more space for more people, more of the time but also by emphasizing ride sharing and pick up and drop off rather than private car ownership you're also working towards that goal of lowering private car ownership and this can only be done through data you need to first be able to manage pick up and drop off zone you need to be able to manage sidewalk used in different ways at different times you need to know who's coming and going out or at least you need to know the volume and capacity of that rather than personal details about anybody but volumes and capacities to be able to plan this kind of infrastructure for buildings innovation again our most exciting facet is not a digital facet but I'm still going to mention it which is that this would be the first district built with mass tall timber all the buildings were planned to be built with tall timber and they'd also be the tallest buildings in the world this is enabled by digital processes it's enabled by designs of a modular kit of parts and a digital fabrication process that would take sourcing of timber through factories and into construction what we're most excited about this is that it results in incredible construction time savings which equates into cost savings which for us can be put back into affordable housing I want to mention the outcome based code quickly but there's a lot of use of environmental sensing for industrial uses and for environmental monitoring but we think that a similar kind of technology can be used within mixed use developments to allow a greater mix of tenants so not just residential commercial but even light industrial kind of mix for a more vibrant sort of live work environment you still have to monitor, you have to know that that's going to be safe and we think that the technology that's available today can help to do that all of the innovations we're doing in buildings as I said are resulting in something we're very proud of which is a commitment to a 40% below market housing this will be two to three times more affordable housing that's been delivered by anybody else on any of the developments in the waterfront in the entire history of the development there then in terms of district infrastructure I would say we're introducing a lot of sort of innovative technology into district infrastructure I mean traditionally most of this kind of infrastructure already works with control systems that has digital layers to it already but there are a number of different advances that we're proposing that we're very excited about and that will significantly change for example with the storm water management we're proposing to pilot some technology that's actually already piloted here in New York with the EP they work with one of the companies that we're interested in working with it adds a layer of smart controls to the system anticipates weather conditions it can flush out a system before a storm so there's better retention levels and keeps storm water out of 90% reduction in storm water into the municipal system through the design that we're proposing at the Keyside site it's a huge differential amount and waste collection speaking about circular economies one of the things we're excited is the implementation of a pneumatic system into the site but more than that it's actually the information about the recycling that's coming through we're working with a local materials recovery and recycling facility to look at the garbage that's coming through to be able to provide information back to residents about how they're doing on recycling we think that an education loop it's really about human activity that education loop back to residents about how they're doing with recycling is the only way that you can really move the bar on improvement to recycling and waste we're targeting an 80% reduction of landfill for our waste collection at Keyside which is radical compared to what they have in Toronto today which I think is somewhere between in multi-tenant buildings they get somewhere between 17 to 23% variably between residential and commercial but one of the other elements we're most excited about is energy systems because achieving climate positive at Keyside is our goal it comes from just great design at the beginning passive house designs even for these larger buildings renewable energy sources Toronto has a pretty green grid which is great but augmenting that with solar and with geothermal but then also one thing we're excited about is the use of AI to manage the building we know looking at buildings today we model buildings in a much more sophisticated way we have ASHRAE standards all sorts of other standards that are relatively different ones but that we use to model buildings and to require for permitting of buildings but we know actually looking at the performance of buildings in true operations the buildings just aren't performing at the level that they're supposed to according to the models so why is that? a lot of that is just the human operations of the building they're simply not able to manage buildings in the way that they're modeled and so we're really excited about the use of building controls that will help automate which will absolutely be able to be managed and controlled by humans but also help to take over some of the functions that aren't being able to be done today to really radically lower the use of energy in buildings so the last one I don't want to miss is talking about social infrastructure in Toronto we're not looking at actually providing services we think that's done effectively by the government and other community groups and organizations in Toronto but we're very excited to partner with those kind of groups there for this we're really thinking about inclusion and equity whether it's providing digital tools where communities can get more involved in their own programming whether it's digital literacy programs or whether it's access to digital technology on the site these are the kind of things that we're very excited to partner with local folks to deliver for the project so the result of all that is a set of outcomes that we're very excited about 40% below market housing as I mentioned 85% reduction in greenhouse gases at the site and a 60% reduction in car trips these are the metrics that we're aiming for and believe we can hit with the project but having said all that we are a private company and what does it mean to implement all of these or propose to implement all of this digital infrastructure I can see some smiles in the front row so one of the other things that I wanted to mention or talk about today is really the framework for how do you govern all that because it is incredibly important and we do think of ourselves as good actors and we think it's incredibly important to address the implications of all of this so we've had extensive consultation in Toronto we've met with more than 20,000 Torontonians face to face in workshops and other activities we've had more than 200,000 people participate in online webinars we've had grant programs summer fellowships and we've had a lot of discussion in the press a lot of discussion in community forums so some of the messages we've been hearing in Toronto are summarized here today so no tech for tech's sake no data for data's sake we shouldn't be gathering or implementing anything that isn't directly tied back to achieving those outcomes that we're aiming for and that the government is interested in we should build in privacy from the beginning of development of any systems and then the outputs whether it's data outputs or whether it's technology that's implemented those should all be seen as resulting in providing results for the public good we've generated at the site especially in public realm or on land that was originally public we see that as an asset that should be a public asset and so we believe that that data should be actually made available to everybody and should not be privatized no single entity whether it's us as a private company another private company or even a government agency should be able to monopolize either that technology or the data that results from it what does that actually mean in practice how do we take those that kind of feedback and think about how it's actually acted on so one thing that we've done is we've developed a set of responsible data use guidelines and an assessment system that we use in-house when we develop all our systems so we look to make sure that whatever we're developing has beneficial purpose that anything any data would be publicly accessible by default we look for transparency in implementation so what are we implementing where is it going to be physically in the environment what is it collecting where does the data go who has governance over that data that should all be clear to the public and to government if we're developing anything with AI how do we develop that responsibly knowing the biases that can go into AI from the very beginning because of the data sets that are used how do we keep data to a minimum keep it secure and keep it de-identified at source and by default and you know just frankly speaking it's been a big I think misunderstanding in Toronto because we are an alphabet company and we're a sister company at Google we're interested in urban development we're not interested in advertising so we don't we're not interested in selling sharing or using personal information for advertising in any way that's just not our business model but it's one thing to say that we're doing that for ourselves but that's certainly not enough we're still a private company so we firmly believe that all of this should be overseen by government so in Toronto we've proposed an urban data trust that would be an independent government entity to oversee to review plans and systems designs before they would be implemented in Toronto it may not end up being an urban data trust one of the things that's been very exciting about the project is that both the city and the provincial government are actually doing a refresh looking at their existing data governance and privacy policies and regulations and they're actually aiming to update those over the next year partly in response to the project which we actually think is a very positive thing because we feel that these the government should be overseeing what's happening for this type of technology another thing that we're thinking about is how do you how do you bake in that transparency from the very beginning of a project so one of the things I'm excited about is that we're thinking about what materials would we submit with a formal development application to the city right now in Toronto, I think it's the same in New York when you hand in your first application for a project you don't have to provide any sort of information on a digital architecture of a site so we're thinking about what kind of system the architecture diagrams diagrams that sort of show the layering of the digital and the physical together what materials could we provide that would make that clear and transparent from the beginning of the project then those would be sort of reviewed and go through iterations through responsible data assessments and overview and then at the end what we think is equally as important is thinking about operational capacity we don't think that all systems and cities should be privatized we think it should be capacity that the government has so how do you work with government partners through co-creation through using regulatory processes to educate and to knowledge share with them how do you do that in a way that you can build capacity on the government side to take over as much as they can of these systems into operations of a site the last thing I wanted to mention was then and just when the project ends we think it's really important that people have visibility over this digital layer which is often invisible to people so we've worked with hundreds of partners to develop an open standard and visual language these are icons that would be displayed anywhere digital technology is implemented they're icons that describe the purpose of the technology the type of equipment that's being used whether the data is identified or de-identified who's responsible for the project or for the equipment and how can you learn more so today maybe if CCTV cameras are being used you might get one small sign that sort of tells you about this this gives you much more information it starts to create a shared language and it gives you a way to actually find out more and to provide feedback so that's a super quick and shallow introduction to the project if you want some light reading our plan for the project is called the Master Innovation and Development Plan it's on our Sidewalk Toronto website it's a mere 1,500 pages with a mere 40 appendices but if you want even more information we will be actually at the end of this month publishing a digital and a specific digital innovation appendix a lot of the appendices are planning appendices there's a lot of detail there already but this is specifically about the digital architecture for the site so that will be on the website it will be public for everybody so please read if you're interested so thank you very much for a very thoughtful presentation the next speaker is Jennifer Ding who actually will compliment this presentation quite well Jennifer is a solutions engineer at NUMINA translating data into insights for NUMINA's partner cities around the world previously she was the founder and CEO of Parkit a computer vision company that developed algorithms to transform internet protocol cameras into a real time parking data sensor, venture backed by Jaguar Land Rover, deployed across North America with major state universities municipalities and acquired by NUMINA in 2018 one of the things that when you look into NUMINA and the work of Jennifer they measure the forms of street level activity with a privacy first approach so I know that many people with this last presentation think about privacy and the mission is to make cities more responsive so they are safer, healthier and more equitable and one thing I really like about looking at Jennifer's background so online and different things and certainly NUMINA there's a lot of discussion about privacy by design intelligence without surveillance and on their website they say they make cities more walkable bikeable, equitable with data so with that said Jennifer please come up Good morning everyone I'm so excited to be here today for this conversation on digital materiality I'm Jennifer Dang from NUMINA today I'll talk a little bit about how we at NUMINA approach building the infrastructure and data platforms for creating street level data in a human centered and privacy first way but first a bit about the past prior to NUMINA so as a born and bred Texan for the first 20 years of my life this was my material environment cars were equivalent to transportation and a mobility challenge was getting stuck in traffic so perhaps it wasn't surprising that when I founded a company our mission was to reduce the pain and energy waste of circling for parking I applied what I was learning as an engineer to develop some computer vision algorithms to collect real time parking data the company was called Parkit and we worked with parking operators like the University of Michigan truck stocks in Oklahoma and we worked with their cameras often times they already had IP cameras for security reasons and we would just turn them into data collection tools so it was a wild three years in parking and by the end I had not only become more aware of the wider world of mobility outside of cars I had come to recognize the growing relationship between city streets and the digital technologies that were moving from the online to the offline world so last year I joined the NUMINA engineering team and soon I realized that just like parking lots where if you want to get data today you usually manually count parking spaces many aspects of our cities are still measured like this as well so traditionally it's a familiar sight so if you want to get pedestrian counts or impact studies traffic engineers or interns will go outside with a clicker and a clipboard or maybe a fancy tablet and as you can imagine this process is slow, inaccurate and inexpensive so often times we don't have good data on how people walk and bike in streets and streets are then planned around what we do have good data on which is cars so in cities like the one I grew up in transportation becomes prohibitive and dangerous for other modes so in joining NUMINA one of the big things I learned was that the same technology computer vision can be shaped to achieve very different goals NUMINA uses a camera as a sensor to measure all modes in streets and right now we focus on five main modes pedestrians, bicycles buses, trucks and cars but we're continually developing new object detectors like wheelchairs, puddles scooters, trash, dogs, you name it if you can see it, we can measure it and we don't just measure it we also track the paths and behaviors of these objects but perhaps the most important part of how we do this is that we do it in a privacy first way all of our imagery is processed on the device, our sensor and we do not save or transmit video the data that we do save and transmit is anonymous movement data so we don't need to collect personally identifiable information we do save a small subset of images a random image per sensor per hour that's used for algorithm validation and training but those are deleted after that process so this differentiates us from other camera based traffic and smart city solutions we are committed to providing intelligence without surveillance as the camera networks expand across our cities we believe it's important that this infrastructure is built with privacy by design to reduce the risk of use cases for the surveillance state surveillance capitalism or targeted hacking processing at the edge is also a much more affordable and scalable approach to urban sensing so here's one of our sensors in Brooklyn our proprietary sensor is purpose built for streets it's designed so that cities can easily deploy it themselves and they usually attach them to light poles with the same steel straps that any street sign uses because we start at the sensor the data creation level we can be the gatekeepers of what data is and isn't collected so we can balance the value of real time data without risking citizen privacy or data misuse so often our users are urban and transportation planners facility managers of the public realm and they access our data through a web dashboard which focuses on volume counts, path visualizations and what we call the numino behavior zones which is our way of spatially filtering data the behavior zones highlight another advantage of using a camera as a sensor instead of putting a tripwire or an array of sensors on each lane or road or sidewalk segment you can draw behavior zones anywhere within the sensor field of view so one numino sensor can do the work of a street full of tripwires we're also developing some new metrics like speeds, dwell time and direction another way people can access our data is through our API which can be used to build new applications or trigger services in the street where needed, when needed the retailer could use our API to get foot traffic alerts on activity in front of a specific storefront or a sanitation department can trigger a trash pickup when trash reaches a certain level on the sidewalk, a common problem here in York or an autonomous vehicle company may want the most up-to-date information on street conditions for the best routing we just released a public API that works with data from downtown Brooklyn so if there's any geospatial data junkies out there that want to play around with our data, please let me know so we're currently deployed in 15 cities in 3 countries and we're about to have our fourth country our first Asian deployment we sell directly to cities but we also have customers in real estate academia, parks museums, business improvement districts so how do cities use our data today so much of my work focuses on transforming our raw movement data into actually interesting insights based on what the cities are the questions of the cities that we work with have and I call this translating data into human so I'll start with some examples of desire lines or ways that people vote with their feet to identify the critical paths that lead to a certain area so here's one of our deployments in Nijmegen in the Netherlands and we color each mode with a distinct color and pedestrians are green bikes are blue, cars red and most of the time the activity we see in this plaza is just really all over the place it's just the sea mostly of green and blue which is always exciting to see but then one day we saw that the paths had turned into this instead and what had happened was a snowstorm as many of the urban planning folks out there know this is a great way to identify the critical paths that people use most often and in this case it happened to be that bikes were mostly going across this horizontal line at the top of the image and pedestrians were carving out these three main roads through the plaza and this data was important for the city so they could prioritize where to put new improvements like a bench and a ladder fountain and to understand where potential mode conflicts were occurring between pedestrians and bikes that were crossing the same area but perhaps a more critical desire line use case happened in Jacksonville, Florida which at the time of our deployment had the highest pedestrian fatality rate in any major American city there would be miles between crosswalks and we saw was that the desire paths really showed crossing activity all over the street but it was really concentrated in this one area so this data could help Jacksonville identify where they should put a crosswalk so that it could match current crossing behavior most closely and now for a local example downtown Brooklyn, last fall we deployed four sensors on the Fulton pedestrian mall we had 100% sensor uptime from November to April about six months of data for 460 million data points of 26 million objects the customer downtown Brooklyn partnership or DBP had a lot of questions that they were interested in understanding what was happening at the street level but one big focus was vision zero metrics so specifically around road safety and traffic violations we were able to look into some of them using our behavior zones so one big question was what is the impact of construction on pedestrian safety during the fall there is really construction all over Brooklyn so this was a big question for them and just for a little bit of background Fulton mall is a two lane road of two way traffic and there is just this constant negotiation between pedestrians, bikes and vehicles for the sharing of the space so when there is a small disruption like a truck unloading for a few minutes or a bigger disruption like months of construction this really impacts how the space is used and shared and this is something of course that DBP knew what had an impact but the real question here was understanding qualitatively what behaviors were occurring and also quantifying what the impact was on safety and as you can see here construction took place on the left side of the road and scaffolding was erected over a significant portion of the sidewalk and it actually shut down part of the sidewalk and as a result the pedestrians which are those green dots had to walk in the street to make their way down the road or cross in the middle of the street so we looked at activity in that yellow zone and what we found is that in the weeks that scaffolding was present pedestrians were 53% more likely in the road and this information could help DBP understand where they might put a temporary crosswalk or better signage in the future when construction does occur and something that I found significant was that this problem really scales up when you consider the city of New York as a whole because at any given time we have over 300 miles of scaffolding another question that they asked was where do cars driving on Fulton Mall come from? This road is actually designated transit only but if you're ever around there you can definitely see private cars driving into Fulton Mall all the time so anecdotally the customer had a theory that these cars were coming from a smaller side street handover place which is the red zone but actually we found that in a week of the 3700 cars that were driving into Fulton Mall were actually coming from Flatbush intersection so this data they can use to prioritize where to put signage and another piece of information we shared with them was when the peak times were for these violations so this they can use to understand when best to deploy enforcement so to conclude here's our team we are passionate about cities and building technology to improve how we live and move in them this year we're working on some exciting projects around trash around Grand Central measuring exhibit engagement at a museum and analyzing mode conflict at intersections we are a mission driven company dedicated to empowering cities with data and enabling the future with an API for streets in this future we imagine that cities are more connected efficient and equitable for all and technology supports rather than inhibits this process to learn more we have some oops some more information on our website on our blog we have a detailed privacy policy explaining our rationale for how we do things the way we do and some case studies from all around the world and if you go to Numida.co backslash API we have our API sandbox there thank you very much Jennifer lots to think about there more questions for you coming down the road the last speaker for this panel this morning is Vincent Lei and he's actually going to compliment the panel quite well because we'll have a discussion around the legal and regulatory frameworks when we think about digital materiality Vincent Lei is a technology equity attorney at the Green Lining Institute where he develops green lining strategy to protect consumer privacy prevent algorithm bias and close the digital divide as an attorney Lei works with the California legislature to pass laws and regulations ensuring that low income communities have the same access to the technologies and tools that are vital to economic opportunity when you look at the work of the Green Lining Institute and certainly with Mr. Lei you'll see that they spend quite a bit of time thinking about consumer privacy they look at the modern and digital form of redlining and looking at uneven access to technology and they look at the role of cities and businesses play in using these technologies whether it be GPS broadband connected sensors cameras and so forth much of what Jennifer Ding just talked about so at this point I'd like to bring Vincent up to the stage Thank you all for having me like Mama said I'm a policy and regulatory attorney at the Green Lining Institute and what I do is advocate for ways to use technology to close the racial wealth gap so the Green Lining was formed about 25 years ago and using an equity lens we work across a lot of different policy areas environment, transportation, technology with the idea of improving social mobility and economic opportunity and this is particularly we focus on areas that are redlined for a little history lesson you probably may all know redlining refers to policies where the government created lending maps for rural red lines around communities on the basis of race banks would evaluate mortgage lending risk whether to give small business loans based on these maps and areas that are red were deemed hazardous and people living there couldn't get loans and this refusal to lend this refusal to lend is really key because access to credit whether to buy a house to start a business, to fix your house or go to school is key to economic equity in the United States you can survive on income but wealth is what provides economic mobility it lets you get financial security it gives something for your kids when you pass so these overtly discriminatory policies lasted for 30 years and from that time that federal government back to $120 billion in loans 98% of that went to white people so if you put that in another way for 30 years white Americans were amassing intergenerational wealth to start homes, buy businesses buy homes, start businesses and send their kids to college and people of color were locked out of that opportunity and that dynamic is the root of today's racial wealth gap so for every dollar a white family has a black family has 8 cents and the disparities are not accidents these have been deliberately caused by government policies so as we think about how to reshape our cities using digital tools and new technologies more data it's important to keep this history in mind so to me the potential of digital urbanism this new way of understanding the ebb and flow of cities using technology and data is key to use that understanding to undo the effects of redlining but before you can do that let's focus on making sure we don't use this data to reinforce and recreate patterns of discrimination and an example I like to use are Amazon's prime delivery service maps a couple years ago Bloomberg did an investigation of where Amazon provides services and this is what the delivery maps look like when you're black you're twice as likely to live in an area where Amazon decided you don't get service so Amazon says it doesn't use race to build these maps and I agree, I believe them but the point is that the legacy of redlining and the geographic aspect of the racial wealth gap means that you don't have to use race data to discriminate against people of color now I know this isn't on the same level as denying bank loans as denying same-day delivery service but the takeaway to me is that some geographer, a data scientist or engineer at one of the largest tech companies sat down and used all of the latest data and tools and managed to make a map that discriminates in the same way that banks did in the 1930s and this is an example of algorithmic bias you can call it data analytics you can call it AI hack-a-list really quickly talk about what are algorithms and where are they used to me, for the purpose of this discussion an algorithm is just a set of a tool or computer model that processes data according to a set of rules whether created by humans or by computers and they can automate complex tasks and then why are they used is that we just have so much data volumes today there's just some examples up there that we have and we want to derive some value some insights as data scientists say so the idea is we can use algorithms to do that work for us and then the insight could be something like let's look at billions of rideshare data from Uber and they can find out, hey women, more than five miles away from home at night with a low battery are 25% more likely to accept a surge fare that's the kind of insights that we're seeing this data being used for and in terms of the data that's being collected that's relevant as we've seen it's cameras, it's urban sensors it's mobility data it's information from tweets and facebook posts, likes, there's a lot of it and they're being used everywhere for predictive policing so if you're not, you get a bank loan whether or not you get a job and what happens from this is data-driven discrimination so we have new ways to analyze and process data to find patterns but we have algorithmic bias at the end of the day because humans create the algorithms humans influence the data that's made and we can see that when we look at financial lending algorithms it turns out even if you don't look at race and if you don't have a person involved online lenders still discriminate against people of color and they may come pay more even when controlling for the same economic factors like income for another quick example let's look at the number of men and women in tech companies can you imagine what will happen if we make an algorithm look at hiring data from, say, Amazon to determine who to hire? Well, we don't really have to guess because Amazon made that tool and they found out, hey it's biased against women and it's easy to see why because for 10 years Amazon was hiring mostly men and the algorithm learned, hey men are the ones that Amazon likes to hire and should promote so we should hire men and the takeaway from that is these tools and data really let well, the analyzing this data and using algorithms you can inherit really undesirable human traits like bias and discrimination and another aspect of this is when you're looking at data you need to make sure that your data sets are representative and an example of this is facial recognition algorithms are really bad at detecting black faces and that's because the data sets that we're using these use just don't have enough black faces in them and that's an oversight that we at Greenlining want to make sure that when we use these technologies to benefit people like that doesn't happen and that gets into algorithmic equity so how to use data and algorithms in ways to reverse the effects of redlining and close the racial wealth gap and in the urban planning and mapping the context here today you know it's really important because the biggest factor affecting economic opportunity and social mobility is the neighborhood you grew up in so if you improve neighborhoods then we can improve and address the legacy of redlining and so how do you do that and you do that by optimizing for equity you need to ask the right questions from your data you need to guide your data use that prioritize equity and one really local example I like to use is the Oakland bike plan when Oakland tried to create new bike lanes back in 2014 they had a lot of community opposition and when they restarted that effort this year they had a very equity guided vision that asked who do we want to build this for what do we want to do what are our goals and one of those is affordability the other one is safety and these are things that came out of the collaborative process and from asking the right questions of the data the next way we can do this is to ensure equitable resource allocation so we can use data tools and policy making maps in ways that identify disadvantaged communities and target those for greater investment and the first way to do that is to have the tools and greenlining sponsored legislation a couple years back where we made a mapping tool that looked at data from pollution sensors, health records, traffic data socioeconomic data sets to determine if a community was disadvantaged and this is what happened we created in effect a redlining map without using race from a legal perspective because you can't do that when you're targeting investments and the next step after identifying these disadvantaged communities is to build an equitable resource allocation mechanism and what we did was we passed laws in California that made 35% of California's $9 billion cap and trade fund 35% of that money is going to communities identified as disadvantaged so the effects of that are that we've had $500 million going to building charging infrastructure in communities that if you just looked at the historical data low income people don't like electric cars so we shouldn't build charging infrastructure there but by building equity and design you build a way to get to a future where low income communities can use electric vehicles they can charge their cars in their neighborhoods so we think that's been really effective to do that and just to contrast the Cal and Viral screen map that mapping tool that we made with the Amazon prime maps in one we have a mapping and data tool and process applied in a way that denies services to communities of color that have been historically redlined and in the other way we use mapping and data tools in a way to drive greater investments and economic opportunities in those same communities to get back to the representative data sets issue there's a big push to use more mobility data from Uber, from Lyft, from rideshare and it's great you can have a lot of insights but if you think about the digital divide you realize that the people who aren't represented in those data sets who much less like to be represented in those data sets are low income people so if you are using that data you're going to prioritize the needs and the patterns of the rich over the poor and that only reinforces redlining if you don't use this data correctly and one big part of this is that we have all this data but there's a temptation to say oh hey we have enough data we don't need to actually go out and do the expensive process of talking to people and that's wrong that's so wrong and the Oakland bike plan I really like this because it was a very equity forward process and they said this about actually talking to Oaklanders and talking solely on quantitative data over the knowledge and experiences of marginalized communities can lead to incomplete decision making and that's true because not everything comes up from the data you can't know about someone's lived experience by looking at an excel spreadsheet luckily there are tools being built to lower the costs of citizen participation to lower the cost of outreach but if you remember the digital divide and smart phones people who don't have internet aren't going to be able to use these tools so you really need to follow up this process with actual listening sessions you need to go to where people are to gather data and the Oakland bike plan did all of this they had very data driven process where they surveyed thousands of people but they also went to supermarkets they went to encampments they really talked to people all over the city to really understand and get feedback on the insights they were getting from their data and I'm not going to touch on this too much because I'm running out of time but there's so many privacy concerns like the typewalk labs said no data for data's sake no technology for technology's sake and I really agree because with all this information you get more precise targeting you can really prey on vulnerable people and you can get really deeper you can start analyzing people psycho-graphic and behavioral results and get really creepy with it and you can see the end result of that in China they're using facial recognition software to repress and limit the religious freedom of their Uyghur minorities so as we think about where to use these tools you know it's important that we don't reduce people the numbers and we don't use these tools in ways that increase redlining and discrimination it's on right so you see that unfortunately the fourth panelist Mimi Scheller is under the weather so at this time I would love for the panelists to come up so my first question is for you as first you're starting with a community or piece of land that didn't really exist in terms of a community but the first practical question is around financing so you're obviously a private entity and you're working with the city and so forth can you talk about just the financing around this type of development and just the basics of it I mean just in terms of how it all came together just who the partners are and what not I mean that's actually something that's sort of underway at the moment because we handed in our proposal just in June so the kind of economics of the site are actually sort of in deep discussion with Waterfront Toronto at the moment but certainly we would be they've requested us and we're happy to work with local development partners as well so we'd have probably other vertical developers working on some of the buildings of the site but you know in some ways it's not unlike just normal real estate transactions where the city itself needs to decide sort of what they're interested in as outcomes for providing that land for private development and what the value of that is to them because certainly developing and this is less about some of the digital elements but just developing that kind of district infrastructure for energy certainly geotech and things like that is very expensive it's more expensive to develop green energy infrastructure than not it's more expensive to provide affordable housing than not so to be honest in terms of the economics it's much like any other development project in that we need to come to a balance between a return on the project and what the city is looking at for the value of the site and what they're looking at in terms of what they'll get as outcomes from the site so what I will say is we won't make money on Keyside it's a first signature project we're more interested in being able to work on the implementation of so many new things and so we're actually the economics don't pence out very well for us but we're we've proposed further development on the next neighborhood as well to kind of balance that return but we're looking at certainly even an aggregate sort of like a below level normal return as a real estate company would. Okay so my next question is probably on everyone's mind is about data and I obviously you know your private firm and if I was to be the devil's advocate here I'm sure you heard this question a million times you know it doesn't seem like it could potentially be a conflict of interest if you're collecting this data how do you also evaluate the project right so we know that these wonderful things come along and you have excellent metrics I've looked at some of the things online but who's doing the monitoring and evaluation and the second piece tied to that is who owns the data and how is that data being used and I'm kind of cheating here because it's not fair but it's also tied to then what happens at some point if through the monitoring and evaluation process we find out that data has been misused or what not what are the repercussions for that so I know it's a three part question and I can walk you through why don't we start just with the monitoring and evaluation so you are the developer you're working with the city certainly and you have a number of metrics and you say here are our goals and they're wonderful goals right so who's making sure is it an outside entity or is it yourself or is there a firewall between that to understand if you're meeting those goals certainly we don't believe it should be us in the proposal to Toronto this really gets to that question of long-term capacity building with government because we shouldn't be self-regulating that but to be able to monitor and measure that government needs capacity to be able to do so so we're actually in that this is one of the big things that we're debating with Waterfront Toronto now because in the proposal we actually proposed sort of a series of new entities I mean they would still be governmental entities but there would be entities that would be stood up to really build up that capacity to be able to have the oversight for what was happening on the project you know they're slightly less comfortable with that which I think is actually a great thing what they're saying is our government agencies that exist today should be able to build up that capacity so let's not think about new entities let's think about which parts within the existing agencies can build up the capacity to be able to do that in the implementation of that practically we also have proposed a number of different tools so for example for energy management we're proposing to build a tool that we would then hand over to the government that would actually be able to sort of aggregate the energy data and show that we're meeting the metrics so there's a need to be able to very practically to address how do you measure and make sure that you can measure if you're proposing these kind of things proposing metrics that we want to be held to was the first question second question was data who owns the data so this is what I touched on briefly in terms of data being a public asset so we don't believe that any entity should be it's certainly no private entity should be able to monopolize the data so in all of the infrastructure designs that we're proposing essentially the data would be there would be data standards that would be signed off with government so literally the standards of how do you organize and provide that data the data would all be provided openly so whether it's an urban data trust or it's a existing government agency some body within an agency they would essentially steward that data that's coming out of all of the different systems would be provided to that organization so then it could be openly used by any other entity whether it's government or other private entities to look at other sort of innovations on top of that data so that we think that's incredibly important and then the third question was what happens if they're wrong I mean that's something we're also sort of actively talking about now obviously you need some sort of practical oversight you need to be able to audit you need to be able to prove that that's not happening and you need to be able to hold organizations accountable for it so I don't have a quick answer for that today because it's something we're sort of actively working on but that's to say that we agree that it's something needs to be actively worked on and you need practical solutions for that fantastic I appreciate it I'll switch to you, I think maybe you will be able to have a conversation here with Narissa certainly in your work you talked about behavior zones and you collect NMI's data and sort of how we can understand our social environments, how we can understand the built environment, the way people go throughout their daily lives can you talk a little bit more about the behavior zones and who are some of the people that have been coming to you to use your data or whoever in terms of partners and whatnot sure yeah besides I guess to address your last question besides our direct customers who are usually city officials so DOT or various other agencies in public agencies we have worked with universities before so we did a program with new lab that was a collaboration with some New York cities and we did some data sharing there I think definitely a big question on our mind is how to best open the data to the communities that we're actually monitoring in because completely agreed this data is on their community it's something that they should have access to so that is kind of what we've been exploring with our API sandbox by opening up that downtown Brooklyn data well let's get that that's like and then the behavior zones yeah I think that goes in pretty well with the data people are interested in different aspects of the scene so for example at one of our deployments in Grand Central right now the partner we're working with is completely interested in detecting trash which is great we're happy to do that but there's a lot of really interesting activity going on in the zone as well for example one of our sensors has a great view of dedicated bus lane so we've had conversations with transit center for example with a lot of the better bus action plan work to see if we can monitor some bus metrics about delays and maybe some violations of cars using the bus lane so I think that there's a lot of opportunity especially in some of our urban environments to make the data and especially the behavior zone data available for more people that might be interested in using it so I think it's a great idea to actually touch a little bit on my next question which was on community engagement how is the public interacting with this data I think you said you were in 15 cities in three countries so how are you using that aggregate data to better inform the work that you do or is there anything that you can say that you see that's cutting across either themes or some kind of similar challenges I know it's a big question but I was just wondering if you could speak to the analysis yet but one big question that does come up especially across our urban environments is mode share versus road share so this idea that how much space do we allocate to cars versus people and how many people are actually using the space so in New York for example I know New York DOT recently did a study where they found the majority of space is definitely allocated towards cars but we have so many pedestrians and cyclists users that are using the space so how can we use the data on the number of commuters of these other modes to make the case that we actually need to make more space on the road for these other modes so it's definitely interesting to compare across New York for example with some of our pilots in the south like Jacksonville or New Orleans where maybe they are a little more car friendly another thing that we're starting to look at is modal interactions so how pedestrians might interact with cars or bikes and something that we know in New York we are all very aggressive commuters so one question is about yielding behavior we are going to be deploying in Portland soon and we know that their driver characteristics are more willing to yield to pedestrians so this is something that we've heard through conversations with the city but something that we'd also love to actually measure quantitatively to see how that might actually affect safety in deeper ways. I wanted to say how excited we are about technologies like NUMA because in all of the use cases we're looking at for streets or for public realm these days with the development of technologies like this which are protecting people's data privacy we can actually implement every single use case except one with private data everything that we're looking at counting the cars coming in for pickup and drop off whether it's that or whether it's park usage it can all be done without collecting any sort of personal data from people and scrubbing it before that data goes anywhere so that's really we think it was a game changer in terms of being able to use tech responsibly in public space great alright why don't I I'm not a data scientist so isn't it possible that you can take that anonymized data and it's really easy to de-identify data these days and you add that to someone's location data from their cell phone maybe because they had Google Maps open and couldn't you identify people and where they travel and put all that together and look at maybe that person doesn't yield a lot and then you get an email from the government saying hey you didn't yield and they take that data with your behavioral data to craft a message in a way that is like most designed to make you feel bad is that possible yeah it is so I mean I guess that's the point and I know everything was shallow and quick but that's the point of having a process like a responsible data use assessment because in that assessment you're committing to what you are and what you're not doing with the data that would be a regulated assessment so if you said you're going to do XYZ with the data you're not doing ABC you're not taking it and then matching with other data sets to identify people and yes then you have to make sure that people aren't breaking what they said they're going to do but it's providing that transparency as to what the data is going to be used for and committing to not use it in other ways so that's why those processes and that oversight is so important because you can do that kind of stuff of course well Vincent why don't you continue along those lines I could just go sit down with a bucket of popcorn and just watch it why am I here no this is great so I mean you've talked a lot about you started out with red lining and certainly it's a very serious issue we know and when we thought about what some of the solutions obviously it was legal and regulatory where it would be the Community and Reinvestment Act or the Fair Housing Laws and so forth and then you know I think about my when I started out with framing today's discussion the time at MIT were talking about the digital divide and who has access to technology and so forth and now you've talked about algorithmic equity which is an important concept so I'm curious about solutions I know you talked quite a bit like the Cal and Viral scan certainly as an example of how you can use technology and information to sort of but when you think but some of the things you presented not only going from the legal and regulatory but also down to the production and the innovation of the technologies themselves so I was curious of how at green lining and the work that you do are you partnering with the private sector whether it be Amazon, whether it be Twitter, Facebook, what have you to kind of address some of the issues that you've discussed right so I think we work a lot on capacity building in government agencies right so we want your regulators your government your legislators to know how to process these data we don't want the regulators just to take companies at their word right so that's one big part of it another part of it is creating the legal framework so California in January is going to have the strictest privacy laws going to effect and we hope that becomes the standard it's kind of based off GDPR in Europe and that gives a lot of people more control over their data and their privacy but in terms of actual solutions you know I think we really need to change a lot of the civil rights laws that we have you know because we have laws in the books that prevent disparate impact right so if you see we work a lot of bankers we look a lot of bank data and getting more transparency in data is really important you know we there was a freedom of information act with banks and insurers and they found out that you know insurance companies were charging minority neighborhoods twice more than twice as much despite having the same driving records right so we need first we need data from these companies and we need the capacity from our regulators to analyze that and that's a really important part but they're also like within companies there are statistical techniques for example racism and redlining in the data but using a lot of them so say you find out your banking algorithm discriminates against people of color you can't legally fix that by saying okay we'll give all these black people all these Latino people a bonus in their algorithmic score that's also illegal because that violates the equal protection clause so we have two really competing laws that say like once you recognize that your data is also racist it's illegal for you to have a race based way of fixing it so you know that's why cowl and viral screen was an important tool race neutral but still generates equity for you know communities of color but I think in a larger scale we really need to rethink our civil rights laws okay so I have a question for all of you and it's tied to the earlier discussions each of you brought this up and when we think about planning development design all those things many of you oftentimes discuss process and I'm curious not only with your partners whether it be government entities the private sector and so forth but what is your process to ensuring if you think about today's society and certainly strong market cities that Toronto and New York Beijing London whatever it is the level of inequities right and you all discuss in many ways the importance of creating either a quality you've mentioned quality of life your communities, economic opportunity educational opportunity all these wonderful things but through the process that you're working on can you talk about how you're actually embedding that in the work that you do through these processes sure I mean I think the affordable housing example I gave is maybe the most obvious one but you know I think a typical developer if they sort of developed the kind of mechanisms to build at these lower rates would simply take that money and it would be you know profit on their bottom line but we don't we really are not interested in building sort of wealthy enclaves that's what's you know that's a huge problem in the last 20 or 30 years in cities as populations have moved back in we're really interested in building communities where everyone can live so you know that's part of the reason our project doesn't pencil out you know for the first one is that we're we're not interested in building something that's not going to have that component to it and so we think by the way we're going to provide as much affordable housing as possible we'll end up with a much more diverse neighborhood and that'll make it a stronger and more interesting and vibrant place I mean it's an interesting question because I mean I guess our work is to get other people to embed equity in their work right so that means that's going to you know the California Senate it's going to the FCC it's going to policymakers we can pressure these decision makers to implement the laws and the rules that we need to kind of force people to consider equity and in their design of things and you know this is part of it right you know it's like teaching people about what are the impacts of all these new technologies and I think it's an iterative process but with enough time we can get the momentum we need to really shift the legal landscape the regulatory landscape so that I design you embed equity into your you know data use Jennifer yeah I think the first kind of challenge that Numina started or hoped to address was to look at equity of vote so we need to you know increase our data sets not on you know car users car owners but also other users of the streets that was a starting point I think probably the biggest procedural decision we've made is that we do focus on working directly with cities so in that way we hope that the groups that we are working with are representing their citizens interests and taking into account the context of what really goes on there so for example I recently heard a presentation by the Oakland DOT about increasing traffic but a big point that they made was just that in many of these neighborhoods access to transit or bikes might not be possible so not only is car use sometimes necessary but it's also a big part of culture and identity so I think just having that local context and understanding it's not like always cars are bad or always transit is the solution but embedding the data in the context of what is going on and how do we decide what metrics to focus on is a big focus of ours fantastic so we have about 10 minutes left and I would love to hear some questions for the audience there's a microphone going around this gentleman here thank you very much I'm Jack I am the founder of something called gizmo which is about data geographic information systems this is a question that I'm going to speak in the 50s and early 60s but there are precedents that help that happen that go back to the 1920s particularly I'm thinking of one the informal steering of realtors by race and color and ethnicity or whatever away from certain sellers and landlords to other sellers I'm thinking of something called restrictive covenants which happened all over this country and in New York City in the 1920s where land was developed and restricted to just white people or restricted against people of certain faiths Jews and Catholics and certainly people of certain colors half of the city of Seattle was laid out in such a way as north Seattle could not in this one is could not be sold to people who were black New York City Jackson Heights garden apartments could many of them could not be sold to people of certain faiths of certain colors and the same thing was true of the co-op apartments on Fifth Avenue Central Park West and Park Avenue that's a precedent for something that creates the block busting any comments on that? Did you study restricted covenants? I think restricted covenants are the other half of redlining you put those two together you get urban blight because in black neighborhoods there was a lot of white owners and those white owners knew my residents, my renters can't move anywhere so why do I need to fix my house to ruin and people who actually own homes in that neighborhood they saw their property values decrease because all the houses around them that were owned by white people and rented out to black people were falling apart so that creates urban blight and then when they created the interstate system city planners are like these areas, these black areas, these redlined areas are blighted they're slums, they're ghettos so let's just build a freeway through them that created a lot of urban issues that we see today suburban sprawl pollution so all of those things go into why the restricted covenants and redlining go into why no one has wealth today in terms of solutions the equality of opportunity project that found that what neighborhood you grew up in is the biggest predictor of social mobility one of their primary solutions is that we need to reintegrate neighborhoods and to re, because restrictive covenants codify segregation we need to reintegrate neighborhoods and I think the challenge today that we all deal with is how to do that without displacement, without all the negative effects of gentrification thank you short question what's the role of the courts especially the federal courts are there any cases yeah so I kind of touched on this earlier there was a supreme court case called RISO that pretty much made it very difficult and if not illegal to correct for algorithmic discrimination on a race based way so that's how the court interpreted the equal protection clause and the disparate impact laws that we have on the books and that's the law of the land until we change the way that we think about disparate impact and equal protection so I'm Rene Sieber I'll be talking later today so I have read all four fourteen hundred years of the MITP my question is it's a call to action so I'm really excited to be here thank you what are we mostly urban planners going to do but my Ph.D. is an urban planning what are we going to do after this conference so much of the time I see the plans of action being outsourced to lawyers as the litigation and laws are the only solution and I don't know maybe it's a structural defect in planning remember I am a planner but you know that we speak to the powers that be so often we're just seduced by the technology and oh the data is out there there's nothing we can do but where is the Gandalf in all of us you shall not pass you know we shall not use this data set because in an era of automated decision making there's no such thing as anonymity anymore excellent question thank you I respond to that the core of the question being what should we how should we act urban planners well I guess that's why I wanted to present the material in the way I did which is to say that you know digital technology and data is potentially inexplicable and so when we think about developing places we have to think about them holistically and some of the most important things you can do are not related to digital things and we have to find a balance of all of that and then we have to do it ethically so I mean personally I think that's what we're trying to do at our firm it's often hard for people to believe because we're a private company but we're a private company of people who come from urban planning backgrounds and governmental backgrounds we're a mission driven company as well and I think we are actually trying to do that by the project itself and the way that we're designing it yeah I mean I think we rely on lawyers it's job security I guess because urban planners you know we're kind of the middle man between the data scientists and the planners and the policy makers and you kind of need groups like greenlining and other lawyers to translate the learnings and the insights and the negative effects of that into effective policy and I wish that wasn't the case and I think of some solutions to do that is like we always advocate for diverse and inclusive teams as a way to see your blind spots because I don't think in a lot of these cases that I presented no one was people went into it with like hey I'm going to go discriminate against black people but there's blind spots hey I'm going to use data sets that only have white faces but when everyone on your team comes from the same background you kind of create blind spots that prevent you from self-regulating yeah as someone in tech definitely very well aware that legislation will be lagging the algorithms, the data sets will be produced and developed and the pace is just not something that can be matched I think legislation definitely has its place and I hope we are seeing a lot of anti-facial recognition legislation happening now but of course there's more than just faces that can be used to identify a specific person right there's a lot of different forms of data and regulating every single one especially if every single city has to make their own legislation just not really something that is going to protect citizens so I guess as a company that does work with planners a lot of the time I think our big ask is if we can find ways to work more closely together because I think for us we think of ourselves as hardware and software providers right but the data we do create and collect should be shaped by both planners and hopefully citizens too so I think there's amount of guidance that should happen in this relationship as a professor it's so hard to be quiet on that question it's a great question I would say think about where change happens and I would say it's a multifaceted approach takes lawyers, community organizers, developers, architects so forth but anyway these are all great responses excellent question audience questions but I do want to ask the panelists one last thing when you think about your work your individual work or just sidewalk labs in general what is the one challenge that you're grappling with the most what keeps you up and thinking about the future and where we're going when we think about digital materiality equity, health all the things that you've laid out for all of you just sort of 30 seconds if you can I know it's a big question and we can start in this order we're always going to you Jennifer why don't we start with you and we can work our way that way if you don't mind I think the expanding camera networks both fixed and mobile is one of the most terrifying things for us so not just CCTV cameras that we already know are all around us but the phones that we carry the invisible sensors that are around us these are increasing and for some kinds of data I think we do have some protection but for others we really don't and so I guess what we're hoping to see more of is a stronger either legislative or even industry focused set of baby coat of ethics or set of standards that can help set a common understanding of what should and shouldn't be done I think maybe what should is a little fuzzy right now but what shouldn't is becoming to be clear what keeps me up at night is that like we don't know what we don't know right like the all this investigative reporting was behind the examples that we found there are whistleblowers you know so that's how we found out that insurance companies charge you know black people twice as much this is how we found out that open data sets how we found out that banks were discriminated against you know people of color so the thing is a lot of data isn't open and it's not free it's not transparent there's no accountability so you know I want to build mechanisms where there's like some sandboxes or regulators can look at the data without you know infringing on trade secrets and you know all these what private companies complain about when we ask for transparency so that they can actually look at and find out what's going on and what's going wrong and you know just even the threat of that can ensure compliance so that's what I hope to do thank you I mean I think my answer would be very similar you know it's amazing when you look at what's already implemented today and as I was saying earlier the kind of invisible layer the invisibility of all of it on the street and out in public spaces and so how can we work to really provide that transparency I mean also for me I think as we what keeps me up at night is as we it's a lot about equity as we think about what we're proposing to develop in this neighborhood there are in the vast majority of the cases there are we can collect data anonymously we can protect people's privacy we have no interest in surveillance cameras but some of the larger infrastructure systems essentially would run the district and so is there truly if you come and live in the district you can choose to live there if you're a wealthy person but you can't if you're going to come and live in affordable housing so that raises real ethical questions of equity and if we run an energy system that has AI built into it you can't necessarily opt out of that and so we're really grappling with what does that mean like could we provide meaningful ways to opt out of that that wouldn't penalize somebody and so to be honest those are the ones I think about a lot because you know I'm happy to say that 95% of what we're doing is going to be anonymous it's like I truly ethically believe there's no issue with it but there are a couple of these use cases where because I'm so focused on environmental protection and that being incredibly important to me the tradeoff well it doesn't have to be the point is it shouldn't be a tradeoff between aiming for climate positive and having systems that can help you get there but protecting people's rights and equity so that excuse me fantastic so with that we'll end this panel discussion but I want to thank Jennifer, Vincent and Noressa for a thought-provoking presentation and discussion and thank you very much and I would encourage you to look at our colleagues work Mimi Scheller who could not be here today because she's under the weather but to see the work that she also does and how well it complements the wonderful work of our colleagues on stage today thank you