Machine-learning traffic analysis piloted in Phnom Penh

For analysing Phnom Penh’s challenging traffic congestion, SMMR works with PPCA and MPWT on novel ways to analyse traffic flow employing smart technology. This week, the two cameras for gathering footage for the pilot became operational.

Despite widely installed traffic surveillance cameras, Phnom Penh gathers its data for planning using manual counting, which tends to be inaccurate. In cooperation with the local company, LUMA System, SMMR developed technology for automatic traffic detection based on machine-learning technology. The technology is customised to detect the specific vehicle fleet of Cambodia, such as vendor vehicles and Cambodian Remorques. Apart from counting, the technology also analyses the speed of each vehicle.

With official permission given by the Phnom Penh Capital Administration (PPCA), two solar-powered CCTV cameras have been installed at different locations along National Road 2 (between Kbal Thnal to Ta Khmau). The cameras will capture footage, to be used for training the model further and validate the findings that ensure the accuracy and reliability of the traffic analysis technology under local traffic conditions.

Results are expected continuously over the next months. Gradually, the technology will become ready for roll-out to other locations, using the existing traffic surveillance cameras. Meanwhile, SMMR will work with the stakeholders on integrating the gathered insights in planning processes.

Grasp some insight into detecting the Indian tuk-tuks on the road. It is work in progress, so watch out for the tuk-tuks that don’t get caught.

Auto-Translate »