Machine learning, artificial intelligence and autonomy are high on the list of future technologies on the horizon, but there are many ways in which these advances are already being put to use. In this session, artificial intelligence and machine learning can be applied geospatial data, increasing efficiency, lowering personnel time, and in some cases, finding what humans might miss.
Session moderated by Thomas Haun, Turner Staffing Group
Examples of Where AI (Deep Learning) Can Add Value to Geospatial Data
We will review the basics of AI including where deep learning fits in, the keys to successfully applying deep learning technology, and the pros and cons of how the industry is offering deep learning technology for you to use. Drawing from our own experience, we will look at examples of how deep learning can be applied to solve problems with geospatial data.
David Jarron and Kevin Miller, Solv3d
Utilizing AI in the GeoSpatial World
Artificial Intelligence, yes AI, is everywhere, all of the time, solving everything whether we want it or not. GIS is no different, but what is reality and how much is hype? What should we reasonably expect from AI today and could it be as revolutionary as we believe in the future? Dr. Aaron Morris will tackle these questions and more with his expert overview of enabling AI in GIS technologies to create real and meaningful learning. Using a recent NYC DoT case study as a real-world example, Dr. Morris will demonstrate how AI can work for city streets. As usage increases and traffic snarls become an accepted phenomenon, there is a greater need for better information about roads, curbs, sidewalks, and all things that exist on and around these assets. By outlining how AI and machine learning decoded and located the state and status of NYC's bus stops, Dr. Morris will show that AI can improve mobility and infrastructure for one of the world's largest cities and how it can work for you.
Aaron Morris, Allvision
AI and Machine Learning from Lidar and Remote Sensed Data
In this presentation we will be discussing the advancement of lidar, lidar Machine Learning and AI and the combination of geospatial information and tabular information to better provide solutions to clients. Currently, this method is mainly being used in the utility sector, it is starting to branch out into other sectors. It is a very powerful tool that isn't utilized to the extent that we are using it in the Utility sector. Additionally, discussions on how this is being used for Emergency Response, Predictive growth modeling, asset inventory, Network Hardening, and Warehouse Crane safety and inspection will be presented. The wrong and then right data thought process method will also be explored during this discussion. Lastly, how this thought process can be applied to all solution sectors will be presented.
James Young, Pointerra US
Speeding Post Disaster Recovery through ML
The ability of government, NGO, and insurance organizations to rapidly assess and deploy resources after a disaster has a dramatic impact on the recovery time for the impacted community. Through our work with the Geospatial Insurance Consortium (GIC), we have developed tools to collect aerial imagery and classify damage using the latest airborne sensor and Machine Learning techniques. This presentation covers the recent developments needed to enable this work, examples of real-world events, use cases, and lessons learned.
Robert Carroll, Vexcel Imaging