Autonomous Perodua Bezza by Reka – artificial intelligence project for software, machine learning

Amidst the slew of new models for the Malaysian market, the 2023 Malaysia Autoshow also hosted an autonomous vehicle test bed for artificial intelligence using a Perodua Bezza, built and operated by Reka, a Malaysian artificial intelligence automation company which specialises in artificial intelligence, software and machine learning for autonomous driving.

The artificial intelligence-driven autonomous vehicle project at Reka commenced in 2016, and has since evolved with advancements in technology and in the industry perception of autonomous technology for industrial purposes, according to Reka.

This exercise in testing autonomous driving in the Perodua Bezza used – as seen in these images – focuses on software and machine learning, says Reka COO Mohd Amri Mohd Din, which is why the modifications carried out for autonomous driving are via additional hardware and actuators affixed to the existing driver controls for the steering, throttle and brakes of the production car.

A sensor suite comprised of cameras, lidar and radar supplies data to the artificial intelligence onboard the test bed’s autonomous driving system in order to identify and understand obstacles, thus helping the car make decisions on navigating a course and how to respond to a range of scenarios by controlling acceleration, steering and braking.

Outside of the walking-pace demonstration held at MAEPS for the 2023 Malaysia Autoshow, the autonomous vehicle test bed is tested exclusively on the dedicated test route for autonomous vehicles in Cyberjaya, and the Reka lab is also situated in Cyberjaya for ease of logistics for the autonomous vehicle testing team, Amri said.

Beyond the potential for self-driving vehicles on public roads such as those being trialled in Cyberjaya, Reka hopes to highlight the potential in autonomous vehicle technology for industrial purposes, said Amri.

“We’ve done a project in Singapore that aimed to increase port efficiency while reducing accidents due to human error. By programming our autonomous vehicles to prioritise safety, we were able to develop a reliable and efficient solution for industrial use,” he said.

Further afield from the automotive realm, further developments by Reka in this area also led to the creation of PLExyz, the firm’s crowd analytics for indoor navigation, the firm said.

This technology employs AI and machine learning algorithms to analyse and predict crowd movement, which Reka says can be useful for crowd control purposes in a variety of settings. The potential for this technology is vast, and could help improve safety and efficiency in a range of industrial and public applications, said Reka.