SBIR-STTR Award

Automated Passenger Counter (APC) has been used in the transit systems for automated passenger counting and reporting. These systems have suffered counting accuracy issues, particularly at high rider load points, and at end-of-line count reconciliation. It was found that APCs have limited use on rail vehicles due to very wide doors. It has also been a challenge for transit agencies to track the origin and destination of riders. Traditionally, tracking the origin and destination of riders has been done using labor intensive and costly origin and destination surveys. Therefore it is desired to have a system that can automatically track riders. In Phase I, Migma Systems propose to develop a standalone APC sensor using stereo camera and laser scanner. To accurately count multiple passengers boarding and alighting transit vehicles, the sensor uses the laser scanner to identify passenger and map them to the corresponding stereo images for minimizing the false detections. To track a particular passenger, this sensor can also extract a set of occlusion–proof biometrics used to find the same boarding and alighting passenger in the same transit vehicle. Moreover the passenger counts, images, biometrics, GPS and transit vehicle ID can be wirelessly transmitted to a remote server.

Automated passenger counter has been used in the transit systems for automated passenger counting and reporting. These systems have suffered counting accuracy issues, particularly at high rider load points, and at end-of-line count reconciliation. In Phase I, we have successfully developed systems for passenger counting and re-identification. Test results using the data collected at a transit bus have shown that Phase I system performed well. The primary goal of Phase II is to develop a commercial product for transit vehicle passenger counting and re-identification under any conditions. It will be a small hardware sensor with embedded microprocessor, and can be easily mounted inside a bus or train, aiming at passenger doors. It draws power from bus or train, counts passengers boarding and alighting, and re-identifies the same passenger who may board and alight at different bus stops or train stations. Sensors have built-in WiFi wireless communication capability. To transmit the passenger data to the transit data management center, bus drivers or train conductors can push one button for data transmission offline. All of the passenger data will be wirelessly streamed to the server and stored. Our biometrics are non-invasive and MBTA will participate in