Working with Inkryptis, Skylark Labs brings computer vision-based passenger counting to public transit, giving operators real-time occupancy data that drives better scheduling, route planning, and rider experience.
Transit agencies plan routes and schedules based on incomplete ridership data. Manual surveys are expensive and infrequent. Infrared beam counters miss passengers who board in groups. Without accurate, continuous occupancy data, agencies cannot match supply to demand -- leading to overcrowded buses on some routes and empty ones on others.
Legacy beam-break sensors miscount when passengers board together or carry large items through the door. Manual ridership surveys happen only a few times per year, leaving agencies blind to seasonal and daily patterns. The absence of real-time occupancy intelligence means transportation operators are making resource allocation decisions on outdated assumptions, degrading both operational efficiency and rider satisfaction.
"You cannot optimize what you cannot measure. Accurate, continuous passenger counting is the foundation of modern transit operations."
Skylark Labs' AI replaces legacy sensors with camera-based counting that tracks individual passengers through boarding and alighting, providing occupancy data accurate to within a single person. Computer vision models built on the Kepler platform track each individual independently, maintaining accuracy even during rush-hour boarding surges.
Real-time occupancy data is available to dispatchers and riders, enabling informed decisions about boarding and routing. Historical counting data feeds predictive models that anticipate ridership by route, time, and day of week. The system also generates NTD-compliant ridership reports automatically, eliminating manual data collection and entry. All processing runs on edge AI hardware installed directly on the bus.
Live passenger count is available to dispatchers and riders through Skylark's Living Intelligence dashboard, enabling informed decisions about boarding and routing. Every data point feeds back into route optimization models that continuously improve scheduling accuracy.
Data-driven scheduling matches bus frequency to actual demand, reducing wait times on busy routes. Precise passenger counts validate fare collection data and identify revenue leakage points. Real-time occupancy information helps passengers choose less crowded buses, improving the overall experience. Right-sizing bus deployment reduces fuel consumption and operating costs while maintaining service quality.
The shift from periodic sampling to continuous measurement transforms how transit agencies allocate resources. Agencies can now validate fare collection data against actual boardings, identify revenue leakage, and demonstrate ridership trends to funding bodies with auditable, timestamped evidence from the Kepler analytics platform.
The Inkryptis partnership brings a level of ridership intelligence that was previously available only to the largest transit systems. Skylark Labs' vision-based counting delivers continuous, accurate data that transforms how agencies plan routes, allocate resources, and serve riders -- setting a new standard for public safety and operational efficiency in urban transit.
See how AI passenger counting can transform your transit operations.
Schedule a Demo
