When the technology required goes beyond road widening and the contract does not seem ready
Free flow has arrived, and not timidly. In just a few years, what used to be a seminar topic has become a contract clause, an agency ordinance, an opinion of the TCU and, more recently, a preliminary court decision. CCR RioSP has become a showcase (and testing ground) for a model in which electronic toll gantries replace physical toll plazas on the BR-101/Rio–Santos highway, within a regulatory sandbox of the ANTT.
On paper, it is a nice story. ANTT itself promotes the project as synonymous with smooth traffic, tariff fairness and environmental benefits, with data on reduced fuel consumption, emissions and operational waste when comparing gantries with traditional toll plazas. States such as Rio Grande do Sul (Block 2) and São Paulo (Rota Mogiana Lot) are already announcing studies and tender notices that include automatic tolls, free flow gantries and intelligent transportation systems as part of the standard investment package.
In the infrastructure narrative, technology appears as just another layer of modernization on top of road widening, third lanes and pedestrian bridges. But when one looks at litigation and regulation, there is a sense that the concession contract still thinks in terms of “toll plaza with a barrier”, while toll collection is moving into the world of data, artificial intelligence and algorithmic enforcement.
The case of the Dutra highway in Guarulhos laid this friction bare. In October 2025, the 6th Federal Court of Guarulhos granted a preliminary injunction in a public civil action brought by the MPF to suspend the fines imposed on drivers who failed to pay the free flow toll on the stretch operated by CCR RioSP. The decision prohibited the Federal Government and ANTT from applying the penalty provided for in the Brazilian Traffic Code (CTB), pointing to disproportionality and defects in the regulatory basis, in the context of an ANTT pilot project within the regulatory sandbox.
The uncomfortable detail for those who design contracts is that ANTT itself argued in court that the fine is an essential part of the enforcement mechanism: without it, the incentive to default increases and the economic and financial sustainability of the system is weakened. The judge, however, saw the user’s conduct as a mere contractual default, not a traffic violation linked to road safety, finding a “misuse of the sanctioning purpose”.
In other words: the contract and the regulation seem to assume that users will always pay, under penalty of a strong sanction; the Judiciary, faced with a new system, reads the same fact as a civil breach, insufficient to justify the severity of the CTB fine. Technological innovation here did not fail in the hardware, but in the institutional coupling: contract, rule and sanction do not fit well together.
The same CCR RioSP faces a different kind of dispute in the Federal Court of Rio de Janeiro, in the public civil action challenging the intra-municipal toll in Mangaratiba under the free flow model. The judgment, issued in November 2024, describes the gantry as the only connection between districts, discusses the lack of alternatives and the impacts on the low-income population, and examines exemptions, discounts for frequent users and tariff fairness parameters provided for in Tender Notice 03/2021 and in the Rio–Santos concession contract.
Note: the Judiciary is not debating whether the sensor works or whether license plate reading is adequate, but what the legitimate design of toll collection is when the toll ceases to be a physical barrier and becomes a virtual filter. Technology, once again, appears less as a technical issue and more as a trigger to rediscuss the very concept of adequate service, proportionality and equal treatment in concessions.
On the external oversight side, the TCU is also examining the “technology + concession” combination, but through another door: that of economic and financial balance. ANTT Technical Note 10537/2024 records that, in case TC 024.813/2017-6, the Court pointed out failures by ANTT in applying the rules defining specifications and prices for ITS System items in the concession, in the context of an audit of tariff reviews and Procrofe investments. Objectively: when the contract reached the stage of discussing rebalancing, rebalancing discounts and ordinary review, it became clear that the reference baseline for ITS items was not properly sized.
And if the yardstick is uncertain, each tariff review becomes an ad hoc debate about the price and scope of technology, with the TCU pointing to the need to improve both ANTT’s procedures and the contract wording.
Meanwhile, ANTT’s own rules are trying to catch up. Recent resolutions of the Highway Concessions Regulation and acts that assign to the Agency’s superintendences the management of intelligent transportation systems and free-flow toll collection make it explicit that ITS and free flow are no longer exceptions and have become a regular subject of regulation.
All this points to a curious scenario: technology is already contracted and regulated as if it were “natural” in concessions, but the institutional environment still reacts to it as something new. ANTT needs a sandbox to test the model; the TCU still identifies flaws in defining ITS specifications and prices; the Judiciary suspends fines that underpinned the system’s enforcement; municipalities challenge intra-municipal gantries on grounds of distributive justice.
Join JOTA’s news channel on WhatsApp and stay on top of the country’s main discussions!
For those working in the sector, the question is perhaps no longer whether we should have ITS or free flow on highways; that seems a given. The question is a different one: have our draft tender notices, risk matrices and rebalancing clauses really been redesigned to deal with innovation, or have they merely fitted gantries and sensors into the same “works and services” logic with which we have always treated additional lanes and resurfacing?
If technology continues to be treated as just another line in the spreadsheet (“ITS System: so much CAPEX, so much OPEX”), it will tend to produce exactly the opposite of what it promises: more uncertainty for the concessionaire, more disputes for the regulator, more litigation for the Judiciary and, ultimately, less predictability for users.
Perhaps the point is to admit that using concessions as a vehicle for public innovation policy in transportation has a cost: it requires more flexible contracts, more responsive regulation and a franker conversation between agencies, oversight bodies and the Judiciary about what is acceptable risk, what is a design flaw and what is simply the price of experimenting with technology in essential infrastructure.
Concessionaires seem to already accept this cost. But what about granting authorities, users, regulatory agencies, oversight bodies and the Judiciary? Do they accept it too?
