In order to join virtual sessions, you must be registered and logged-in(Were you registered for the in-person meeting in Denver? if yes, just log in.) 
Note: All session times are in Mountain Daylight Time.

GIS 9000 | 2019: A GeoSpatial Odyssey

Authors: Matthew Tenney*, University of Toronto
Topics: Geographic Information Science and Systems, Urban Geography, Geographic Information Science and Systems
Keywords: AI, GIScience, Big Data, Urban, Spatial, Ehtics, EhticalAI, Artificial Intelligence
Session Type: Paper
Presentation File: No File Uploaded



Artificial Intelligence (AI) has reemerged, once again, as the next geospatial frontier in urban analysis. IT seems that computational resources have finally caught up to the ambitions of deep learning and convolutional networks, and the opportunities are genuinely aspiring for the applied uses of AI in a variety of urban applications. Using examples from the city of Toronto we present how AI can alleviate a vast number of laborious tasks and improve our overall efforts in serving citizens. These include the uses of large crowdsourced datasets to learn urban perceptions from street-level imagery. We also take a much more pragmatic look by using semantic image classification to identify urban infrastructure and assess accessibility for pedestrian safety. We even demonstrate the potential of AI in “taming” the usability of LiDAR towards the ends of green-roof and storm-water analysis. In many of these applications, it seems the GIS 9000 is responding with a confident: “I am completely operational, and all my circuits are functioning perfectly.”

However, it has become apparent that the realities of being artificially intelligent do not entirely match the expectations as advertised, or always serve a real citizen. We present our viewpoints on the potential of AI while discussing the real barriers of its implementation, the risks of its use, the interests at play, and ultimately the responsibility we have to know when to say: “I’m afraid I can’t let you do that.”

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login