 Our next speaker is Jerry Scott, Scotty, and he's got some, he currently works at the Naval Postgraduate School and the Information Sciences Department. I'll let him give you a little bit more detail about what his expertise is, but I'd like to have him and welcome. Great, thank you. You can hear me all right. I'll walk around a little bit, but I'll try to stay up here. So I'd like to have this this talk a little bit different, particularly with a group of this size, just interrupt me as we go through. I'd love to just turn this into a conversation. So a couple of things that I seem to run into a lot right up front, so I'll answer a couple questions before we even get started. The first one is choice of terms, why I use the word autistic instead of person with autism or person with ASD, autism spectrum disorder. That's deliberate, that's based on the policy statement from the autistic self-advocacy network about identity, particular autistic identity being so central to a person that a person with doesn't seem to apply very well. And so I use that, I do kind of go back and forth a little bit because I bounce between academia and hacking communities and sometimes you have to say what needs to be said. But I do take that approach, I take a neurological diversity approach to my research and try to be inclusive of autistics in my research, not studying that thing, but studying with and growing with. The second thing that I often get is why does the Navy care about this? I'm at the Naval Postgraduate School and I'll get to that in the workplace a little bit, why we care about autistics in the workplace just like any other information-centric business. But we also have through the Congressional Mandated Research Program, a deliberate autism research program in the military studying both adults and children as they relate to families. My work is primarily based on adults and I'll talk about my daughter here in just a little bit. But that's kind of the questions up front that I do like to address. So the the ARSIS lab there, Applied Research and Cognitive Information Sciences, we are just standing that lab up at the school. It will be a wet lab, it's a cognitive psychology based, the lab director, Dr. Malamick Weyer, a cognitive psychologist out of let's say Claremont graduate, was her PhD. She's a newest faculty member and it's a shift in direction for the information sciences department to recognize that the cognitive sciences are just as important to what it is we're doing in that field. And so it is I'm the more technology side of it and she's the the more cognitive psychology side of it. But new lab that we're standing up we're getting into some pretty interesting work I think. So why do we care? What we found over the past several years is a recognition and a change in business from treating autism as a disability and disorder to often something that's desired. There are skills within the autistic way of thinking that are useful, particular in the information fields, information related fields. So even in the Israeli military, they've set up a special unit for autistic analysts that they're looking at and you see lots of articles like this. And yet unemployment of autistic adults is one of the highest in the world. This is for the associate for advancing futures, let's see adults, I can't remember anything in the organization now. That's why, because I have it wrong, advancing futures for adults with autism.org. This is their numbers but we are looking at incredibly high unemployment and we see that also in autistic autobiography, which I'll talk about in just a moment. The motivation then is we are not creating the workplaces we need to keep people on the spectrum employed. And that's a failure on the employer side and societal side more than it is on the autistic side. It's not, we need to shift our mindset from adapting the person to the work environment to adapting the work environment to the person. And we have a lot of experience in this. We know how to do this, particularly for physical disabilities. We've been doing it for years and years and it's gotten much better. So the picture you see on the on your left there is from the 1960s. That's not exactly a conducive work environment for somebody who has to work out of a wheelchair. Leaning over a desk, working off his briefcase, reaching over phones. And we've come to the point now, a military guy, where we're sending soldiers with missing limbs back out to Iraq and Afghanistan. Because we know how to address physical disabilities. There's lots to do in that area as well. Excuse me, there'll be, and there'll be a couple presentations on that this afternoon. But we have a 50 year head start. People working with access for physical disabilities has kind of a 50 year head start, if you will. Another problem that we find is that autistics often have a challenge of social interaction. And so self-advocacy becomes an issue. If you are, if you have a lot of social anxiety and are too nervous to address the problem of your environment with your boss, or you don't quite know how to say the lights are too loud and the noise is too bright and things like that, then you can't create the place where you need to work. And that brings up an associated problem is that there's so much variation within autism. People on the spectrum in the room, so I'm sure there's some of us, and I extend the spectrum because if we take a diversity approach to this, autism is not a thing. It is not just this one disorder that we can now address. On a side project, we're working on what I'm calling a multi-dimensional description of neurological difference. Because everybody's just a little bit different. I am not diagnosed with autism, I am diagnosed with ADHD, but I have significant sensory issues. And so I create my office environment in a way that allows me to work and focus. But I don't have other things that are commonly associated with autism. And so we need to be able to define and address that diversity of experience in a dynamic way. We can standardize, excuse me, see if I trip over everything here, we know how to design environments that are prosthetic in lots of areas. We come up with design principles. The picture in the center is the headquarters of Lighthouse for the Blind and is designed that way to ease movement through the space. You know how sidewalks are these days. With crosswalks that make noise, light up, have thresholds on them, we build environmental aspects to help us access our environment. I'm proposing that we do the same for autism, but it has to look different. It's not going to look necessarily like this. So my concept I've been working on is technology-enabled prosthetic environments. What I realized is that I've been working on this for the last 18 years. This is my daughter Anna. She's 20 years old now, finishing up adult transition program. She's severely autistic, minimally verbal. So how do we create a work environment for her? What is her skill set, if you will? Under the normal path, she's folding towels somewhere. Currently she's folding towels in the fire department, but I think that's because she likes the fireman, not so much the folding towels part. But I've been creating prosthetic environments for her whole life. So you see that's a picture of her really having a good time at a restaurant. She's wearing a hat because she has touch sensitivity issues, so she always, always wears a hat. She's wearing noise canceling headphones because sounds can be too distracting for her. She creates a soundscape with her phone, either it's music or some other YouTube videos or things like that that she's listening to. Her headphones are also connected to my phone, so that if I need to talk to her, I can interrupt with an intercom and say, Anna, we need to go, and it's not so disruptive. And she's typically wearing sunglasses because things are just too bright. And that's stemming from that, and I realize that I keep plugging away at getting technology to work for her in all of these different ways. We come to the idea of the prosthetic environment. How do we dynamically adapt the environment of an individual to fit what it is they need at that time from a neurological perspective? And so the idea of, I guess, TP, I haven't even figured out how to pronounce my own acronym, but the idea is to improve the autonomy and self-determination of autistics. And so the significant issue for someone like my daughter, who right now requires 24-7 constant care, has a habit of wandering off, and she has seizures as well. And so if we're going to enable someone like her to have more autonomy and self-determination, we have to start interpreting her state and helping her to communicate her needs when she may not have the ability to verbalize them. And so we want to look at that in living conditions, reducing caregiver requirements and caregiver fatigue, and setting up work environments that are conducive both to the autistic worker and the more neurotypically worker, more neurotypical worker. If you've worked with people on the spectrum, you can know that sometimes they can be annoying. Of course, a lot of other people can be annoying, too. So sometimes we like to turn down the volume on all of our co-workers. But we need to create the environments that are conducive. And that's what we're trying to do here. So because I feel I'm a little neurological diverse, but I don't necessarily have that experience. And the experience of autistic is different than the experience of the next autistic down, if you will. It's not something that we're going to clearly define and put in boxes. And we're really trying to push against that just labeling a very simple label like I gave, severely autistic and minimally verbal. That doesn't give us enough information. So the study that, I'm sorry. So what we're trying to do with that is to create the technology around the person that senses the state of the person, the state of the environment, and then adapts the environment as needed. So that includes wearables, embedded, portable computing, and a large part of what we're doing is up on the top, this integrating of the technology. That in fact is our biggest challenge. Things like Fitbits and other wearable things often exist within their own silo. And we need that data other places. So we're looking on the computer science side of how we create wrappers around commercial grade technology so that we can incorporate it into such an environment. And this is where I like things to shift to a discussion as we get a little further in this because that's what I like the hacker community to work on. Corporations won't necessarily do this. This is potentially not a very money making scheme here. But if we can begin to integrate pieces of technology better, we can get better use out of it. And so wrapping technology in other software that helps control it is a large part of what I'm looking at. The next challenge is to monitor both the individual physiology and the current environment that they're in. And I'll talk about a study we're just starting off in a few minutes about that. And then we need to allow the individual to have more control to adjust that environment and also support that with automated control of that environment. If for example the environment senses that the sensors on the person see that the environment's getting louder and there's certain sounds that are incredibly annoying to the individual, then can you change the soundscape that they've got in their earphones to offset that? Turn on the noise canceling or change what's being played? It's that type of reaction and reactivity of the system that we're trying to build in and automate. But the study has got to start with the autistic individual. And so centered there is the person and all of the different types of things and pieces of information that we would like to integrate and move back and forth. And I'll come back to that in just a minute. But that's where the challenge is and that's where I hope the hacker community can start helping us plug some of these things together. So got to start with the autistic. So the current study is a qualitative study of autistic blogs of better trying to better understand the autistic experience and define it in a way that technologists who might not share that experience can start to work towards assisting with that. So recognizing that even though these blogs are public and publicly accessible, there is a sensitivity with the individuals in there sometimes. There are access issues. There are some discrimination issues. And so for now we are not associating what we're finding in this review of the blogs with any individuals. So some of you in this room may have written some of the quotes that will come up in just a minute. And we're working through that now with some autistic collaborators on this project of really getting at bringing in autistic autobiography into the discussion of autistic diversity and access and the discrimination that's often associated with that. And so we're trying to give voice to people on the spectrum. And I'm going to rip this microphone off at some point. So I won't go into the details of how we're setting up the study. We're doing this as a qualitative study and what we're finding some things. So these are some early observations of what is being reflected by multiple authors of autistic blogs. And many people are familiar with some more full autobiographies or formal autobiographies. Timberl Grandin's, of course, written a couple books. You may know the book Edo and Autism Land is a formerly nonverbal autistic who now types. This extends past that a little bit into this less formal autobiography. What did people write on a daily basis? What are the posts on Facebook? What goes on Twitter? That's the kind of stuff we're looking at. So first of all, it's a recurring discussion of the physiological phenomena of the experience of autism that is different than what neurotypicals often report. So here's an example. All the different streams seem to individually assault my sensory system and cut to the heart of my ability to function. We have sensory issues completely stopping the ability to work. So how do we deal with the sensory issues and not say, well, then you can't work. So there's often a big distinction between what a person thinks about themselves and what they're doing and what's going on in the external world. Wouldn't say that's necessarily unique to autism, but the expressions of that become very different, particularly when you find a difficulty in self-advocacy in expressing what the changes are. And so if we can reflect that better, we can improve outcomes, I think. Recognition of difference. So we often depict or the autistics are often depicted as being socially unaware and not knowing about their own differences in behaviors or social settings. And what we're generally finding is no, that's not true. They're hyper aware. Just not necessarily as good at expressing it and don't see the need to follow that stupid social rule. And so figuring out how we as society can accept that type of diversity, that's on us. Communication between neurologically diverse and neurotypicals is often an issue. I don't know how to talk about this or I see it from this side, but I create this misunderstanding. And then what are relationships like between the neurologically diverse and the neurotypical? My daughter I wish was more verbal. I would like to know how she expresses or how she views the relationship, for example, with her sister. Which is very close and her sister advocates for her a lot, but I can't get it in her words. And so part of the study is kind of seeing where the experiences are the same in a divergent group of people. But this is where the kicker that's come in a lot is the response to stress is a challenge for many people on the spectrum. The physiological changes may be different. We're not sure about that, but stress impacts almost every area of life is what we're seeing all the time. It increases OCD type behaviors. So one thing that in the neurological diversity movement we want to talk about is the difference between disorder and diversity. There are clearly disorders associated with autism. Epilepsy for one is very common. That's a disorder. If we don't talk about autism as a disorder, we still have a need to address some of these types of things. OCD is another one. My daughter is very strong OCD tendencies that get worse when stress goes up. We can observe that. Let's see. There we go. This multiple streams of noise. Some people have the ability to focus in the noisy environment. Other people don't. That's an issue that I have. There's a conversation going on in that room right now that I can almost hear and that distracts what it is I'm trying to talk to you about. People can be the problem. We all know people that are the problem, but if you can't work in an open office environment, you can't work in an open office environment. That's just how it is for some people. Again, instead of shoving the people into the norms that we've already created, why can't we adjust the environment? And then inappropriate response. This is a social issue. You get stressed. You get silly when you should get serious. How do we let other people know, well, that's okay. And then this idea of how different am I really and how do we address that? So those are the things we're finding. What we're hoping to do with this is use that as the starting point for designing these types of systems. And so where we are right now because stress is so important. And I'll pause just for a moment. So the study we're kicking off right now is to use wearable sensors, fit bit type devices. The one we're using here is a research device, but it has the same sensors we have in our fit bits. Can I use that and other information that I received from the person to in real time identify the onset of acute stress? We believe, yes, we're validating some of these sensors. We're working on some machine learning algorithms. The study that we'll do, it will be fun in my view because we get to stress people out intentionally, of putting wearable sensors on people, sending the data into our algorithms and then stressing them out. And then because we've stood up a wet lab now, comparing that against salivary cortisol, which is kind of the gold standard, if you will, of physiological response to stress. So can we get consumer grade wearables to give us accurate enough information that our algorithms can detect when that stress starts? And then can we do something about it? So these are two elements of my daughter's environment. The noise canceling headphones, the bottom one is a vest. That one is, I can't remember the name of that vest. It has air bladders in it though. And so it's controlled via cell phone so she can get a hug without having to interact with a person. And so the integration step next is to use the data coming from that watch to change the soundscape and turn on the vest automatically without user control. Combined with that is the environmental sensors that we can embed in the home, in the work environment to figure out what it might be about the environment. So one thing that took us about three months to figure out in one school that she was in is that the fluorescent lights that they had in that room were really a problem. And all it took to make her situation that much better in that classroom was to leave the lights off. Very, very simple things like that. And that's where the machine learning comes in on this because if I'm sensing all of these things about the environment and I'm measuring the physiological response to the individual, I can start to record that and associate this type of environmental condition causes this type of response. And so you saw the chart I had up before of all the things we'd like to do, this is what we're working on right now. Can I take heart rate data? Those are more familiar with that heart rate variability in particular and there's several different ways to measure that. Can I use skin temperature, gavanic skin response, movement, right? So is the person being very, very jittery, doing lots of self-stimulatory behavior? Is that an indicator of stress or not? All of the environmentals and in this case just doing two things. Give her a hug and change the volume of her surroundings. Simple things but there's a lot of integration that we have to work on. And that's kind of where I'm at and I really just want to open this up now for a discussion of what might be possible, what might we do, what is it that you're working on that we can plug into this type of environment? Go ahead. So I'm curious about you mentioned the sound scapes. Yes. So she kind of puts together various sound scapes throughout the day. What are your thoughts around kind of relating sound scape data changes in state as related to what kind of things that they're listening to? Yeah, that's absolutely stuff that I want to be able to track more and I do. So one of the things you might, maybe we can hear this. Let's see, I'll play a quick sound scape here if I can. So this is one that we use for focus, if I can turn it up. Nope, where's my speaker? No, nothing. All right, that's a lot of that guys. Anyway, there we are monitoring sound scapes. What I'm not able to do right now, one of the biggest challenges I have, now it starts, I see, okay. If you guys want to ask questions, we want to make sure because we are big, if you can take it up to the microphone so we can hear the question with the answer, that would be awesome. I appreciate in the dialogue it's not always easy to do that, but if we can do that then we can hear both. Thank you. So for example, my daughter will not wear that wristband sensor. No way, no how, never wear anything on a wrist ever. And somebody at UT though is working on a graphene based essentially a tattoo that is the same sensor. That's the type of stuff I'm trying to gather now is who's working on what that is going to make this type of stuff doable. Another question, please go ahead. First, a great talk, can you hear me? I'm not sure it's turned on, is that one on there? It's on, okay, got it. Two things, one just an observation, it seems like this, do you call it TP? Yeah, I don't know what to call it yet. If you've got a better name, I'll take it. It seems like you could use this also for something like depression, just monitoring someone's physical experience and adjusting it like that. Absolutely. Second is a question, do you know how much brain mapping has been done on autism and to incorporate brain-computer interaction with this kind of environment? Two very great points, thank you. So the first one, absolutely, we are very closely looking at this, particularly for PTSD, another thing the military is very interested in, but in discussions with people, the idea of this type of dynamic environmental control we think can help with a lot of mental health challenges, diversity, and people in general. Apparently this room, the air conditioning doesn't work great in this room. At what point does that start to distract all of us from the discussions we're having? And then your second one real quick again was... Right, so there is several studies that ongoing and there's been lots of work in this, but there's a couple challenges with kind of the academic or formal research side of what's going on in there. First of all, to do things like functional MRI requires a certain level of compliance and that can be a challenge in this community. Can we actually get them to work inside of a functional MRI? If you've been in an MRI, they're not quiet and can we make this stuff happen? New things, new tools are coming into use right now that are very useful for this. FNIRs, if you're familiar with that, functional infrared so we can place just an infrared sensor in a certain location. The other challenge or problem that I see in that area of research is that not necessarily where the money comes from but what the money's been allocated to. It's often allocated primarily to diagnosis and early intervention and so looking at where the problems are rather than which parts of the brain are functioning and how they're functioning. A related topic is what we've seen when we do sensory substitution research in several areas. Some research that's going on at Baylor, now going on at Stanford, has to do with substituting what is normally visual information into soma sensory or haptic. What we find in that, particularly in individuals who are blind or if we are doing sensory substitution from audio to haptic in depth, is that the parts of the brain associated with the replicated function are often activated. So if I'm providing visual information by sound, then we see the visual cortex often lighting up and processing that information as a visual way. So there's a lot we don't know on how that works and that will definitely help us to create better environments for this. So thank you for the comments. Yes, please. Hey, thank you. So a couple of things. One, I want to empower you that you have said that maybe this isn't going to be something that can make money but right now the big wave in especially government grant funding, et cetera, is work. So how do we train people and how do we help people work? So I think you can access a lot of grant funding, especially from the NSF, NIH, et cetera, partnering with human computer interaction labs because when you're saying the funding for these medical programs are on prevention, early intervention, but human computer interaction labs are designing for accessibility and usability. Exactly. Thank you for that. So creating those partnerships and then just kind of another design challenge that can lower the barrier to entry for building these technologies. Taking a machine learning approach is going to require a lot of sophistication, but you could also take an approach where you create two finite sets, right? So you have a finite set of stimuli and you have a finite set of reaction. And then what you can do is have this user control. So say, okay, I have this finite set of stimuli and this is the set of reactions that I want mapped to each. And then for instance, in order to dynamically update as they learn and grow, you can have something where once a month it triggers the user to review their controls and see if they want to remap. Yeah, good point. Thank you for that. We are working with the Watson open source community to make some of the machine learning stuff we need to do more accessible, but that idea of user control of giving the person using the environment the ability to manipulate the settings and then have that environment use that feedback is very important. Thank you for that. Yeah, exactly. Other comments? How are we doing on time here? Yeah, we got a few more minutes. Yeah, one thing I wanted to call out. I'm really happy to see somebody taking such a applied take on wearables. I'm really waiting for a couple of years to see those really take on a more meaningful approach that I put LEDs on my body. But really I wanted to circle back on, you talked about your multi dimensional analysis of differences. I think you called it. I'm wondering if you're considering something like a Myers-Briggs spectrum where you have like a series of just a short AB selections that help us a pretty wide variety and are pretty widely known of personality traits and applying a similar model to just at a glance know what potential triggers and stimuli are for these kind of people. Yeah, it's I think it's going somewhat in that direction to identify so things a dimension might be sensitivity to sound and that might but also have to be broken down into volume or certain tone or waveforms and so detangling all of those if you will. So if you take this, if you take every one of these you know around the 360 degrees as a different dimension you come up with a radar plot if you will of how sensitive a certain dimension is and then you can say something like we do with Myers-Briggs and INTP or something like that and so you can start to better identify the differences within the neurological challenges and not just oh they're autistic right I think that is very limiting and how we define it but so thank you. The file has self-disclosure, self-identity. Right. I think there's a degree of advertising that perhaps if not explicitly at least non-verbally so that people approaching know how to deal with these other interactions that are not going to be typical inter-human interactions. Yeah, exactly. In some cases, not every case but in some cases bad with LEDs on it might be a very very good thing when it's connected to my physiological sensors and things like that and it can give my co-workers the red light to say don't talk to me right now come back when I'm in a better shape things like that. I just hang it outside my door in my office and just say don't bother me ever but that's yes absolutely those are the things that we need to work on thank you for that. Thank you. And I think we're about done. I'm going to work the door next hour so if I'd happy please stop by and let's chat more because the next thing that we're doing I'm going back to a website that we had a few years ago and standing that back up called autistic.org to try to get more grassroots maker space projects into there. There's a few projects up there now I don't think you can get to the website right now but if you've got a project that you think might fit into this type of construct I'd love to hear about it love to have you write about it we're working on a internet connected toilet right now because my daughter and many others likes to play in water and likes to see what will and won't flush and that causes me a couple hundred dollars a month plumbing bills and so not letting her flush when she shouldn't that's a design challenge right so thank you thanks everybody love to talk to you more outside