Using Amazon Rekognition to improve bicycle safety

This article demonstrates a machine learning solution using Amazon Rekognition to improve bicycle safety by automatically detecting and documenting close encounters with vehicles during cycling.

• Addresses bicycle safety concerns by analyzing ride videos
• Uses Amazon Rekognition to detect vehicles in recorded bicycle rides
• Automatically identifies vehicles passing too close to cyclists
• Generates video clips of dangerous encounters for potential law enforcement review
• Implements a serverless workflow using AWS services like Lambda, Step Functions, and MediaConvert

The solution allows cyclists to automatically find and document potentially dangerous traffic encounters, helping to promote road safety and potentially enforce 3-feet traffic laws.