FireTracker
Research
My 2018 master's thesis at the University of Bergen examined how an indoor-positioning application could support instructors and firefighters during smoke-diving training. I designed and built the React interface that turned beacon data into a reviewable route over a floor plan.
Separate theses, shared research
I worked on FireTracker from January 2018 and finished writing in November. The thesis was formally published in December. It was part of iComPAss, a collaboration between NORCE and the University of Bergen, and the research phase with the fire department was shared with another student. We wrote and submitted separate theses.
I was responsible for the React web application and jointly developed the Android tracking application. The other student developed the back end. My interface consumed the processed location data from that service, while the mobile application recorded and uploaded data from the BLE beacons and phone sensors.
Reviewing an exercise that is difficult to observe
Instructors cannot follow smoke divers through a smoke-filled building, which limits what they can see during an exercise. The research question was how an indoor-positioning application could support instructors and firefighters in smoke-diving training. FireTracker recorded movement during the exercise so the route, time use and activity could be discussed afterwards.
Designing the review workflow
The web application allowed an instructor to create an exercise, upload a floor plan, select and position the beacons, and open a completed exercise. The review combined the route with a list of recorded locations, timestamps, duration, walking and head movement. Instructors could play the locations as a sequence or move through them one at a time.
The interface developed through four iterations. Early wireframes were reviewed with the fire department, a functional prototype was tested by its training leader, and four information-science students carried out a heuristic evaluation. Their findings led me to simplify the exercise setup, improve its behaviour on an iPad, add contextual instructions and create a beacon-management module.
What the field trial showed
The final field trial involved two smoke divers, two instructors and four exercise sessions. The instructors found the system easy to use, but the phone sometimes received a stronger signal from a beacon that was farther away. The reconstructed routes therefore did not represent the firefighters' actual movements precisely enough for dependable training feedback.
A useful workflow with limited positioning data
The project demonstrated a viable workflow for preparing an exercise and reviewing recorded movement. It also showed that a clear interface cannot compensate for unreliable source data. The BLE positioning needed further technical work before the visualisation could provide the pedagogical value the firefighters expected.
Individual master's thesis by Edvard Pires Bjørgen, University of Bergen, 2018. The research phase and system development involved another student, with testing by Sotra Fire and Rescue Service. Interface figures are extracted from the thesis.