Words By: Drone Major
Since the year 2000, 15 people have been killed by sharks off the West Coast of Australia, and many more surfers and swimmers have been injured.
However, it’s possible that drone technology means that’s about to change.
This autumn has seen the rollout of Shark Spotter ©.
Shark Spotter © uses battery-powered drones to identify sharks up to a depth of 30 feet below water. Live video feed is sent back to the drone operator who, using identification software, is able to identify the sharks in real time.
The software uses a sophisticated Deep Learning Framework (artificial intelligence) for shark detection and recognition. The system’s algorithm is dependent on information collated from videos of sharks available publicly to help it ‘learn’ to differentiate sharks from other marine animals, surfers and swimmers.
The use of identification software increases the accuracy of shark identification in real time, enabling operators to identify sharks with around 90% accuracy. That’s an increase of around 70% of the accuracy of identification achievable without AI software.
Once a shark has been identified, notifications can be sent to the smartwatches of swimmers and surfers.
The technology has been developed by University of Technology, Sydney, in collaboration with the Little Ripper Group, a commercial drone company; and opens up a wealth of opportunity for improving safety at some of the world’s most notorious beaches.