Wednesday, May 27, 2015

Au Québec: A good place to go missing




On May 22nd, 2015 I had the pleasure of attending a meeting hosted by the Sureté du Québec and Université Laval CRG. The topic was using GIS for missing person search operations. We introduced ourselves and quickly began sharing how each agency is currently using mapping for planning and operations. There were over 30 people in attendance from multiple agencies and the faculty at Université Laval. 


  
Our hosts for the meeting, Alain Croteau and Mir Abolfazl Mostafavi

Here is a quick overview of what was discussed.

Software

It seems many teams were using Touratech for a long time but realized they needed to expand their capabilities because they could not load up-to-date base data into this package. They have since begun integrating with ArcGIS Desktop loaded up with minimum essential datasets.

Here is a short list of the software agencies are currently using.

Workflows 

Here are the most commonly discussed workflows.
  • Printing Maps
  • Displaying Basemaps and Base Data (Doris Poulin, Sgt Surete de Quebec had a very impressive minimum essential dataset all stored locally using ArcGIS Explorer Desktop).
  • Editing Incident Data (Initial Planning Point, Clues, Sectors, Assignments, Search Area)
  • Downloading GPS Tracks

Spatial Analysis

This, for me, was the most interesting part of the discussion. Here are two projects that were discussed.

Estimating probability of area for missing persons with Alzheimer’s (Khaled Belhassine, Alain Croteau, & Mir Abolfazl Mostafavi)
  • A GIS model that will produce a grid in which each intersection (decision node) will be given a variable value based on factual, personal, geographical and statistical information correlated by scientific knowledge. 
  • This tool is still being tested but the preliminary results are promising.
Theoretical Search Area based on missing person profile and travel speed (Miguel Blanco)
  • A GIS model that will produce isochrones for missing persons based on their profile or expected behavior. 
  • Miguel is near completion of his MS Thesis and will be comparing techniques to Doherty et al. 2014 


Challenges

Despite great advances in the use of GIS in recent years, the group discussed some remaining challenges that I think we can all relate to.
  • Sharing information between teams. They are only having meetings like this every few years and currently do not have a discussion portal.
  • Training. While GIS software is getting easier to use, without specialists on each team, it is hard to get personnel trained and ready for incident response. 
  • Internet Access. While 3G is becoming more widely available, teams still need to have all of the local data stored as minimum essential datasets and cannot yet rely on having an internet connection due to dead zones across the Province.
  • Standards. They face the same common challenge of using multiple coordinate systems on incidents without necessarily having the capability of converting them “on the fly”.
  • Emerging Technology. Advances in technology bring new tools but also new challenges. How do we adapt search planning and operations to live GPS tracking, UAV / drones, and even augmented reality?

Solutions

I think there are simple ways we can help the Province with some of their challenges.
Also, here is a link to my presentation: http://arcg.is/1dwm0xM 

Special thanks to Pierre and Guillaume from Esri Canada for being there to answer questions and remind the agencies that their volunteers can get access to Esri software through the Esri Canada Non-Profit Organization Program

Overall, I am extremely impressed with the widespread use of GIS for search operations in Québec. I am especially excited about the partnership forged between the Université de Laval and Sureté du Québec. I hope that all in attendance know they are welcome to post questions to the SARGIS Group and will continue to share their techniques with the rest of the community. 


Merci à Alix pour la traduction

Isochrones produces by Miguel Blanco's travel cost model