Whale vessel collision risk decision support prototype
An estimated 80 whales are killed by ship strikes off the U.S. West Coast each year during peak whale season, showing the scale of a problem that is often difficult to detect before a collision occurs.
I built the final working prototype for AI Marine Sentinel, a Python based decision-support system designed to identify areas of elevated whale-vessel encounter risk in the San Francisco Bay Area. The system combines live vessel activity with whale reports and environmental context, then presents areas of higher risk through an interactive map.
After shipping stakeholders raised concerns about rerouting vessels, I revised the approach by proposing AIS-based speed monitoring inside whale-risk zones and implemented that feature in the prototype.
Python backend
Interactive map/interface
AIS vessel data integration
Whale and environmental risk inputs
Risk zone visualization
Vessel speed monitoring inside higher-risk areas
Presented at DigitalOcean Week as part of the project’s development.
Allow time for demo to load (30 sec)
Please open demo in safari