Now Reading
AI-enabled program designed by DOST promises contactless apprehension of traffic violators

AI-enabled program designed by DOST promises contactless apprehension of traffic violators

Tessa R. Salazar

We’ve witnessed this more than once as law-abiding motorists: We see another vehicle in blatant violation of traffic rules and regulations, and all the while traffic enforcers on site stand there as if nothing happened. If they do choose to apprehend, the violator flashes a badge, or drops a name, or sweet talks his or her way out, and he or she is let go, with a smile and a salute even.

Doesn’t that make our blood boil?

Like it or not, these are the unavoidable quirks of contact apprehension, especially in our culture where we take almost everything personally. Laws are distant and unrelatable, and seem to have no immediate consequence to our present circumstance and immediate future, unlike the person in front of us. Thus, the impersonal and objective application of the law becomes subservient to the outcome of one’s direct interaction with the other. And there are so many examples of this on our streets. And admit it, we ourselves may have been guilty of excusing our way out of a ticket one time or another.

Well, here’s some good news (or bad, depending on how you’ve been conducting yourself on the road). If things go as planned, the Department of Science and Technology (DOST) may soon be able to remove all that “personal” factor in traffic enforcement, and let that proverbial axe fall where it may.

On June 24, DOST launched several programs and technologies in support of the Artificial Intelligence National Roadmap to integrate AI in almost all sectors. Among those is the so-called Catch-All. And as the name implies, this AI-enabled program would function as a system of “contactless apprehension of traffic violators” via a “vision-based artificial intelligence analytics software for traffic and transport applications.”

Developed by De La Salle University, the Catch-All system is designed to detect all types of traffic violations and the vehicles that committed them via a smart camera video capture. The system can then apply its detection, tracking, profiling, plate localization and plate character recognition capabilities for a more detailed identification of the vehicle.
The P11.6-million funded project under DOST’s research and development program uses AI in analyzing CCTV-based videos in traffic management agencies. It would be able to monitor, detect and generate reports of traffic violations. Human operators can then take care of validating and filing the reports. The system can help in eliminating such “quirks” in human interpersonal interactions during actual contact apprehensions—foremost among them corruption, bribery, coercion and other forms of “negotiations” between violator and enforcer.

See Also

The Catch-All program, DOST reveals, is about to roll into real-world application. DOST Undersecretary for R&D Rowena Cristina L. Guevara told this writer: “Catch-All is in the commercialization stage. It belongs to the Fastrac (Funding Assistance for Spinoff and Translation of Research in Advancing Commercialization) intervention program for DOST-PCIEERD (Philippine Council for Industry, Energy and Emerging Technology Research and Development) supported projects to help them commercialize technologies produced by our research projects.”

One may ask, how does Catch-All differ from the existing non-contact apprehension system of the Metro Manila Development Authority (MMDA) that also uses CCTVs installed on strategic locations? DOST said that the current system still relied on humans who would need to monitor rows upon rows of TV screens in order to spot violations. Apart from being exhausting for the operators, such a system can be inefficient, as human eyes may miss other violations when monitoring simultaneous events.

On the other hand, under the watchful, untiring, all-encompassing and impartial eyes of Catch-All’s AI, no violation would be left unnoticed.