- Vision software
Every vehicle's class, verified at the toll
Tollsense compares, in real time, the class detected by vision with the class billed by the toll system and flags every mismatch with visual evidence - on 1,200 lanes in production.
- 1,200lanes in production
- 99%+accuracy, independent audit
- 0sensors to add in the lane
What Tollsense does
Vision-based classification
Silhouette, axles, size - the class is determined without contact, on the lane camera.
Transaction cross-check
Every passage is compared with the class billed by the toll system.
Visual evidence
Every mismatch is documented by a time-stamped video sequence, usable for dispute management.
Image quality assessment (IQA)
Application-driven image quality estimation, designed exclusively for tolling: each camera is qualified for the task it serves.
Accessible vehicles
Detection of vehicles for people with reduced mobility and assignment of the correct tariff class.
Presence detection
Confirms that a vehicle is actually in the lane and discards requests issued by an empty lane after a malfunction.
How it works
Capture
The existing lane camera films the passage.
Classify
The vision engine determines the class in real time.
Cross-check
Any mismatch with the billed class is flagged, with evidence attached.
Designed to be integrated
Open system
Existing cameras, off-the-shelf VMS, no equipment lock-in.
Stable APIs
Versioned components, documented releases, rollback by design.
Gated deployment
Factory acceptance, pilot site, field rollout - a written methodology.
Shared O&M
24/7 monitoring, shared operator / maintainer dashboard.
Start a Tollsense pilot
A defined scope, written performance criteria, an exit clause with no residual cost. That is how we start every project.