Tech behind NCAP prevents possibility of profiling and corruption

During the continuation of oral arguments opposing the controversial No Contact Apprehension Policy (NCAP) held at the Supreme Court on Jan. 24, an SC justice questioned if the system operating NCAP can profile motorists and make them susceptible to targeting.

Supreme Court Justice Marvic Leonen asked Atty. Juman Paa, one of the petitioners, if he felt he was being “profiled” by the NCAP system after being tagged four times because of the same violation in the same intersection in Manila.

Leonen said the NCAP system could have biases toward specific vehicle types that have previously committed offenses because of the presence of artificial intelligence (AI) and big data. Profiling could lead to extortion and cause and unjust and oppressive “wealth transfer,” the justice added. 

How the NCAP system works

Can NCAP system profile and target specific vehicles? First, one must understand how the NCAP system works to understand the concern raised during the SC oral arguments. 

Under NCAP, high-resolution cameras are strategically installed in locations designated by local government units (LGUs). 

The system replicates and programs the lanes, intersection box, pedestrian crossing, and stop bars seen by the cameras. Then, the system creates detection lines or zones that define vehicular flow on lanes and spots where vehicles should stop when a traffic light turns red. 

Under NCAP, the system automatically detects traffic violations once traffic footage is encrypted and transmitted and processes the data using AI.

How AI works in the NCAP system

Based on a demonstration conducted by QPax Traffic Systems Inc. during the launch of NCAP in Valenzuela City in September 2019, the system flags a vehicle for possible violation. 

If a car continues to move past the detection lines or zones after the traffic light has turned red, the vehicle is flagged passing through a red light or beating the red light. The system captures all vehicles violating the rules set by the LGU regardless of vehicle type or class. 

Images of the vehicle at the moment of committing the violation and its plate number or conduction sticker is then taken and cross-referenced with the data of registered vehicles from the Land Transportation Office (LTO). This is to identify the vehicle owner, and a notice of violation is sent to the registered address.

NCAP cameras can also detect other common traffic violations like obstruction of pedestrian lane and yellow box, illegally changing lanes, among others. These traffic violations rely heavily on observing road markings and traffic signals.

The NCAP system cannot flag vehicles unless caught committing the violations. In addition, LGUs review the footages to confirm if a violation was indeed committed. 

Profiling and targeting improbable

Under the NCAP system, cameras are positioned to see only the rear side of vehicles and not the face of the driver. The system has no way of identifying the vehicle driver during capture and only depends on data from LTO.

Following the NCAP protocol, the system eliminates the possibility of corruption. Corrupt traffic enforcers in the field cannot easily flag motorists for supposed violations whenever they want to. 

Regardless of the social status or financial capability of the driver or owner of the vehicle, the NCAP system cannot flag a violation that is not committed. With photo and video proof, motorists can do away with the hassle of arguing with traffic enforcers on the street.

YOU MAY ALSO LIKE