AMAG and Bellevue Use AI to Improve Pedestrian Safety
Advanced Mobility Analytics Group’s (AMAG) AI-driven SMART products were used to improve the safety of pedestrians in Bellevue, Washington. AMAG technology has been proactively used to confirm the positive safety benefits of High Visibility crosswalks and Leading Pedestrian Intervals, helping Bellevue achieve Vision Zero objectives. The SMART platform is powered by the AWS cloud. AMAG has created a serverless architecture using Amazon S3, AWS Lambdas and Amazon DynamoDB. Parallel batch processing of data is orchestrated with AWS Batch and computer-vision at the edge is provided by AWS Panorama.
AMAG and Bellevue Study Shows High Visibility Crosswalks Significantly Improve Pedestrian Safety
Advanced Mobility Analytics Group (AMAG) and the City of Bellevue partnered with Amazon Web Services, Ouster, Outsight, Blue-Band, SCATS, and Fehr & Peers to conduct a before-and-after evaluation on the safety impacts of high visibility crosswalks using video analytics. The study found that high visibility crosswalks reduce vehicle-pedestrian conflicts by 56% (see Evaluation of High Visibility Crosswalks Using Video Analytics | AMAG | Fehr & Peers).
Bellevue Expands Use of Leading Pedestrian Intervals After Study Shows 42% Reduction in Vehicle-Pedestrian Conflicts
Advanced Mobility Analytics Group (AMAG), the City of Bellevue, Microsoft, and Jacobs conducted a study to evaluate the safety impacts of leading pedestrian intervals (LPIs). The study found that LPIs reduced vehicle-pedestrian conflicts by 42% at 10 intersections in Bellevue. As a result of these positive findings, Bellevue is expanding the use of LPIs throughout its downtown area.
The insights needed to deliver Vision Zero
The only viable pathway to Vision Zero is to proactively identify high crash locations prior to crashes actually happening. Advanced Mobility Analytics Group is committed to helping its partners achieve their Vision Zero objectives. AMAG’s platform uses computer vision and deep learning AI to analyse video and sensor data to measure, track, and log critical conflicts–known and validated precursors of crashes.
