What did the evidence measure?

8,388 simulated travel around the 311 active points of the country, with a GPS point every 5, 15 or 30 seconds and a 5 to 12 meters error that is dragged like that of a real team: real passes for each toll and 4,674 routes that pass less than 60 meters without crossing it - through the caltiere, in the opposite sense, up or down - half with the GPS running towards the toll and many faster than normal.

Coves on routes that did not cross the toll

0

Actual past charges

74.1 %

Real past that were left as a warning

24.1 %

Actual passes without registration

1.8 %

With equipment that only report position, without speed or direction: 68.1 % of the actual past charges, 28.6 % on alert and 0 cobros on routes that did not cross, in 6,516 travel. In the face of doubt, the alert engine: never lower the threshold to charge more.

And the most difficult test: 4,614 trips that reach the toll and turn around without crossing it - in the middle of the square, in a return or in a link, half after waiting up to three minutes -, 9,024 that stop in the caltiere next to the portico and 4,968 Rear passes with a tail in front of the shed, on each gate of each point and with the two types of equipment: 0 cobros to whom he did not cross.