Engineers at the University of Glasgow in Scotland have developed an artificial intelligence (AI)-based drone system that can assist in search and rescue efforts for hikers lost in the wilderness.

The researchers suggest that applying AI to drones currently used in search and rescue missions — thanks to their higher vantage points that allow for covering more land more quickly than searches conducted on foot — could improve recovery outcomes.

To make this determination, the team created an AI model using data sets that show paths taken by people who were lost and then found by search parties around the world. The team also added data about factors including age, reason for hiking, whether they were alone, walking, on horseback or using some other type of transport.

Geographical information was then added to the model regarding both the path taken by those who were lost and where they were subsequently found. The model was run millions of times, with each scenario representing a simulated search, until it eventually filtered down the most probable paths a lost hiker was likely to take. The system would then instruct the drones to search those paths first.

During trials of the technology, the team determined that the AI model outperformed traditional approaches for finding actual hikers. The new approach reportedly found lost hikers 19% of the time versus the 8% to 12% of the time of traditional approaches.

An article detailing the model, “Deep Reinforcement Learning for Time-Critical Wilderness Search and Rescue Using Drones,” appears in the journal arXiv.