 I hope you found time to view the two videos we selected as preludes to this lesson. They are chapters 2 and 3 of the second episode of the Geospatial Revolution video series by Penn State Public Media. In chapter 2, Powering Business, we learn how United Parcel Service, that 58 billion-dollar technology company that just happens to have trucks, uses geospatial technology to increase the efficiency and safety of its package delivery fleet. Chapter 3, Finding a Healthy Future, shows how socioeconomic and health data were combined in maps that helped underserved communities in Philadelphia gain access to more nutritious foods. Both are examples of how spatial analysis can make a difference for the better in everyday life. Have you studied spatial analysis? You will if you're pursuing Penn State's Master of GIS degree since Geography 586, Geographic Information Analysis, is a required course. That course is based on this book by David O'Sullivan and David Unwin. O'Sullivan and Unwin explain that patterns on maps provide clues to what processes might have caused the patterns. One purpose of spatial analysis, they say, is to determine whether a hypothesized process may have generated an observed pattern. That's a little like reverse engineering a pattern, to understand what factors may have caused it. Spatial analysis helps us answer the question, why is this place the way it is? And by appreciating patterns and processes, we also become better able to model and understand future scenarios. This book covers multiple aspects of spatial data analysis, including spatial data manipulation, descriptive and exploratory analysis techniques, spatial statistics, and spatial modeling. O'Sullivan and Unwin point out that while data manipulation and basic analytics are commonly included in GIS, spatial statistics and modeling tend to be less well developed. That's why so much spatial analysis is done in statistics and modeling platforms like R, SAS, and others. Because of this, your ability to move with agility between GIS and other analysis environments is valuable. O'Sullivan and Unwin discuss the problem of spatial representation, contrasting the object view of the world as a collection of point, line, and area features with the so-called field view of a world whose properties vary continuously in space. And they take pains to consider the pitfalls of spatial data analysis, including spatial autocorrelation, the modifiable aerial unit problem, and scale and edge effects. The book was written with specialized courses for senior undergraduate and postgraduate students in mind. It's noteworthy for the accessible way in which it presents challenging concepts. Still, spatial analysis remains an esoteric topic for many. That's a pity, since most of us on the fortunate side of the digital divide use spatial analysis nearly every day. In this lesson, we're going to focus on a few of the most common spatial analysis methods. Call them everyday spatial analysis. For example, the video about UPS highlights two of the most common uses of geospatial technologies, location tracking and routing. As Doug Newcomb points out in Wired Magazine, believe it or not, in-car navigation systems have been around for over a century. One measure of how widely used vehicle navigation technology is today appeared in a 2017 announcement by data provider here that its digital maps were installed in 100 million vehicles around the world. As you may know, modern personal navigation systems like Google Maps, Apple Maps, and Waze display traffic and accident warnings derived from the tracked locations of drivers using their apps and from the alerts that drivers post. In addition to that crowdsourced information, these systems have begun to incorporate authoritative data from government agencies about road closures due to construction and special events, among other things. GIS plays a key behind-the-scenes role in data sharing since governments maintain much of their data in geodatabases. Uber launched its movement service to help policymakers develop data-driven transportation plans. Uber's terms of use specify that it owns and retains the right to resell the location data generated by its drivers and app users. The economic impact of this technology may be surprising. In 2013, Google commissioned an economic impact study by a firm called OXERA. OXERA estimated that geoservices, the amalgam of satellite receivers and manufacturing, electronic maps, satellite navigation and imagery, and location-based search, generates $150 to $270 billion per year in revenue. Geoservices is a more consumer and business-oriented conception of our field than, say, the U.S. Department of Labor's geospatial technology industry. Even so, the overlap is considerable. OXERA estimated that geoservices save over a billion hours of travel time and three and a half billion liters of gasoline per year. They even reckoned that faster emergency response saved over 150 lives per year in England alone. Route optimization is a topic of long-standing interest and study in the field of operations research. The earliest analytic solution to the shortest distance between two nodes in a network, such as a road network, is said to be Dijkstra's algorithm of 1956. The algorithm minimizes the cost of travel by calculating the shortest distance between an origin and every other node in the network, including the destination node. Of course, to determine the fastest driving distance between an origin and destination, your personal navigation app should also take into account traffic conditions, speed limits, one-way streets, tollways, and other factors beside distance. Route optimization becomes more complex when a vehicle needs to make more than one stop. The analytic solution to the so-called traveling salesman problem finds the shortest route to a series of destinations while visiting each destination only once. The vehicle routing problem provides a generalized solution for determining an optimal set of routes for a fleet of delivery vehicles. The VRP was first presented in 1959 by mathematician and operations researcher George Danzig and colleague John Ramser. Their algorithm and subsequent refinements reflect an object view of the world in which road networks are represented by linear features whose distances and other attributes, such as one-way streets, are known. The vehicle routing problem has since been implemented as a network analysis procedure in GIS, as discussed in the UPS video. Considering the benefits of saved time, fuel, and increased safety, a logical next step is development of autonomous and semi-autonomous cars and trucks. These benefits, combined with economic incentives for trucking companies, ride-hailing services, and related interests, make it a virtual certainty that driverless vehicles will be part of our future. Depending on where you live, the future may already have arrived. One reason autonomous vehicles are so fascinating is that they combine so many geospatial technologies. Google's fleet of driverless cars rely on mobile LiDAR and radar, video, GPS and inertial measurement, high-resolution HD digital maps, and network analysis capabilities like routing. But some worry about the ethical issues that autonomous vehicles raise. Consider the scenario posed by philosopher Eric Schwitzkabel. You and your daughter are riding in a driverless car along the Pacific Coast Highway. The autonomous vehicle rounds a corner and detects a crosswalk full of children. It breaks, but your lane is unexpectedly full of sand from a recent rock slide. It can't get traction. Your car does some calculations. If it continues breaking, there's a 90% chance that it will kill at least three children. Should it save them by steering you and your daughter off the cliff? Similar ethical and safety concerns accompany the prospect of drones used to deliver packages. However, examples of ethically responsible drone use abound. The country of Rwanda, for example, has launched a national drone delivery service to get blood and medicine to patients in remote villages. Closer to home, Penn State MGIS student John Schaefer showed how unmanned aerial vehicles can be used to discourage elephant poachers by monitoring watering holes in Kenya. Geospatial technologies and spatial analysis are relevant to pedestrians, too. For instance, an app called WeWalk combines location tracking and analysis procedure called geofencing and gamification to motivate kids to be active out of doors in San Antonio, Texas. A sixth grader named Estrella Hernandez imagined the app as she wondered how to help family members and neighbors who struggle with nutrition and diabetes. Since then, while still a high school student, she was named Chief Inspiration Officer at WeWalk Games, the startup that created the app. Speaking of nutrition, the second Geospatial Revolution video we invited you to watch is the chapter titled Finding a Healthy Future. It describes how the PA Fresh Food Financing Initiative worked to eliminate food deserts in Philadelphia neighborhoods and shows how GIS was used to perform a classic site selection analysis to determine the best locations for new supermarkets. I won't dwell on site selection here because you'll have a chance to perform such an analysis yourself later in this lesson. Suffice it to say that a key concept in site selection is topology. The set of spatial relationships then includes connectivity, adjacency, and containment of objects. Next, I'd like to consider another commonplace spatial analysis method that the video implied, but did not discuss directly, geocoding. As viewers of this presentation are sure to know, geocoding is the process of turning descriptive locational data, such as a postal address or a named place, into an absolute geographic reference. That's how geocoding expert Dan Goldberg puts it. Geocoding is one of the most commonly used spatial analysis procedures because demographers, crime analysts, public health researchers, market analysts, and others often need to map and analyze address reference data. Geocoding is a fundamental component of spatial analysis that consists of four elements. An input, typically but not necessarily postal addresses. Second, a reference data set. In other words, a detailed digital geography. Third, a processing algorithm that seeks to locate the input data on the digital geography. And finally, an output, such as latitude and longitude coordinates or polygons within which point locations are found. We can assume that much of the data used in the supermarket site selection analysis, including household population, income, and commuting behaviors, were collected for the US census. The Census Bureau counts population and other things by household, but it needs geocoding to associate those counts with voting districts. That's how representatives in Congress are apportioned after all. Another consideration is the fact that although census data are collected about individuals and households, they're reported by census enumeration areas, like blocks, block groups, tracts, and such, to protect confidentiality. For the same reason, sensitive health data associated with householders is aggregated to census areas or other districts. This diagram illustrates the hierarchy of enumeration areas or geographic entities by which the Census Bureau reports its findings. The GIS analysts that performed the supermarket site suitability analysis used geocoded socioeconomic and health data that were aggregated to block and block group levels. Have you used GIS to geocode address reference data? Or perhaps an online geocoding service, like this one hosted by Dan Goldberg's shop at Texas A&M. You do geocoding every time you enter an address into your phone and get back a map with a push pin that represents your destination. Whether you use GIS, an online software service, or just a map app, you should know that geocoding is fraught with errors. As Goldberg points out, analysts and modelers need to be aware of how inevitable geocoding errors may affect their results. So do unwitting drivers who follow directions blindly. Meanwhile, market intelligence firms, like Claritas, compile address-referenced information about consumer behaviors into lifestyle segments that enable marketers to target potential customers. Even Esri has gotten into the act with its own tapestry segmentation data, which you can sample using its ZIP Lookup web app. Check it out and find out what your ZIP code says about you. The segment that describes my wife and me, Xurbanites, is surprisingly revealing. How about you? This case study has focused on routing and geocoding as two everyday uses of spatial analysis. Of course, you don't need a GIS or even a computer to think spatially. Spatial thinking is part of our everyday lives, from judging distances at changing speeds while we drive or cycle to work, to choosing and navigating to the most desirable available parking spot, to placing yourself in the spatial arrangement of a meeting room, to deciphering a photograph or information graphic in news or social media, to playing a game of chess or Minecraft with your kids at home. In this delightful book, Diana, Sinton, and friends point out both the pervasiveness and the importance of spatial thinking. They point out that GIS has the potential to extend our spatial thinking abilities by visualizing patterns, measuring distances, calculating dimensions, and analyzing spatial patterns and relationships. Diana and friends are also quick to caution that GIS use doesn't automatically make people better spatial thinkers. We've all heard about how drivers who over-rely on in-car navigation sometimes find themselves in bad situations, like these tourists who followed instructions to drive into a bay during low tide. However, Sinton and friends go on, valuable spatial thinking may develop if the map reader begins to ask the why questions, such as why are the patterns the way they are? This recalls of Sullivan's and Unwin's point that patterns on maps provide clues to possible causal processes. At their best then, both spatial thinking and analysis are about why, and go a long way toward explaining why GIS matters every day.