SARNET started with a problem. Two problems, in fact.
The first problem to solve was the lack of access to reliable, efficient distress signalling for backcountry users. Current options were leaving consumers with a choice between cost, coverage, and effectiveness. Through my reserach and personal experience in marine SAR, I knew that the lines of communication between a subject and rescue team could be complex, and dispatching an appropriate response could take up to hours. This is unacceptable, given that the rescue team's time-on-scene is the most significant predictor of positive outcomes on SAR calls.
To solve this, I proposed a three-part system, using a networked system of handheld devices and repeaters in a determined area (e.g. national/provincial park) to provide live updates to a central command station. The user device would be a small, simple handheld communicator. It would reliably broadcast a user's location at given intervals, and send a distress signal when needed. It uses low-power, low-cost hardware, and could optionally be paired to a cell phone for added functionality. It would reduce the up-front cost to the user by 90%, eliminate subscription fees, and could be rented or loaned by park services as a visitor safety program. The handheld devices would connect to a redundant network of solar-powered repeater stations, each equipped with antennas to send and receive position and distress signals. These repeaters would pass the information to a central command station, where park staff could monitor the position of each user, and be notified in seconds to a distress alert.
The second proiblem to solve was the inadequacy of backcountry forecasting, specifically in alpine environments. Current numerical weather models are not designed for use in unpredictable alpine environments, and obtaining verification data to improve models is challenging, as the locations are remote and far from weather stations. This poses challenges for both backcountry users and SAR mission planners, as both currently rely on unreliable forecasts to make safety-critical risk assessment decisions.
I propose to use low-cost weather data collection stations at each of the aforementioned repeater stations to gather real-time, hyper-local weather data across a park. This data, in conjunction with numerical model estimates, would be used to train a machine learning algorithm to provide accurate and timely weather updates to backcountry users and SAR dispatchers. This would significantly increase safety in alpine environments, preventing users from encountering dangerous weather situations, and allow SAR teams to use the most adequate resources to undertake rescues.
This project was a great opportunity to combine my passion and experience in SAR with my engineering design skills. I also had the opportunity to explore business model development, proposing a business plan and market strategy to launch this as a real venture in parks across the country.
I am actively developing this project, both prototyping devices and exploring venture opportunities.
