Who passes, how much passes and at what speed — with up to 1% average error.
Classified counts of vehicles, pedestrians and intersections, intercept O/D and stated preference surveys, and speed and travel time studies. In the counts, independent triple checking: AI classification on the video, manual review by axles and type (DNIT standard) and a supervisor's sample check.
The Traffic Surveys pillar measures the traffic on the road itself: how much passes and of which type (AADT, turning movements), where it comes from and where it goes (intercept and cordon O/D), what it is willing to pay (stated preference) and at what speed it travels.
The method is the same one refined over decades: video recording and independent triple checking, with up to 1% average error — the figure that underpins tariffs, impact studies and speed regulation.
Counts & Classification
AADT, pedestrians and turning movements at intersections.
See the service →Intercept O/D & SP
Intercept, cordon and stated preference for toll studies, fares and through traffic.
See the service →Speed & Travel Time
Spot speed and floating car for level of service and speed regulation.
See the service →Controlled field collection
Filming throughout the collection period for later processing: a safe and auditable approach.
Independent triple checking
Three checks bring the average error down to 1%: AI classification in the video pre-processing, manual checking by axles and vehicle type (DNIT standard) and a final sample check by the supervisor.
Deliverables for design
AADT, movement matrices and series ready for capacity analysis and microsimulation.
Related projects
BTG Pactual — Rota da Celulose
Classified volumetric counts with continuous filming per point and identification of the number of axles and lifted axles.
Do you need AADT, an intercept O/D or a speed study?
Counts with triple checking and up to 1% average error. Talk to a specialist.