Grab, Southeast Asia’s leading superapp, in collaboration with Autonomous A2Z (A2Z), a Korean full-stack autonomous vehicle technology manufacturer, announced Tuesday their first autonomous electric shuttle bus pilot in Singapore.

Grab said in a statement that the pilot aims to serve first- and last-mile commutes for Grab employees, connecting Grab’s one-north office with the nearby one-north Mass Rapid Transit (MRT) station.

The electric shuttle bus, which is retrofitted with A2Z’s autonomous software and hardware alongside Grab’s Internet of Things (IoT) devices, will commence operations tomorrow. A trained safety driver will be present onboard at all times to ensure passenger safety and smooth operations.

“Safety is Grab’s top priority, and we are committed to taking a cautious and calibrated approach to autonomous technology pilots,

“Through this initiative, we are evaluating the safety performance of autonomous technologies and exploring the potential new job opportunities this sector could create,” said Yee Wee Tang, Regional Head of Operations, Grab.

Ryan (Ho Jin) Kim, Managing Director – Business and International, A2Z, said this pilot marks A2Z’s first deployment of autonomous technology in Singapore.

He said the firm has been collaborating closely with Grab, the Land Transport Authority, and local safety drivers to prepare the autonomous shuttle for Singapore’s unique transport environment.

“The pilot will enable us to better assess the commercial viability of autonomous shuttles in addressing first- and last-mile commute challenges while exploring their potential to alleviate manpower shortages in the public transport sector,

“A2Z and Grab look forward to sharing our insights and findings with the government to support Singapore’s broader autonomous vehicle ambitions,” he added.

The autonomous shuttle has undergone over 100 hours of rigorous training on the fixed 3.9km route connecting Grab’s office and the MRT station.

The training process includes data acquisition, data annotation, algorithm training and on-route testing.

The data acquisition collects information such as road infrastructure, traffic lights, lane markings, and detecting pedestrians, vehicles and obstacles, while the data annotation labels objects with attributes such as size, speed and direction.

Meanwhile, the algorithm training trains the shuttle to proactively respond to real-world scenarios, such as stopping at zebra crossings, traffic lights, and fine-tuning driving controls for a smooth ride experience.

The on-route testing, on the other hand, optimizes navigation and adapting to unplanned scenarios such as traffic obstruction.

Equipped with 11 sensors, including LiDAR, radar and cameras, the autonomous shuttle has a 360-degree view of its surroundings, enabling it to detect and safely manoeuvre around new objects and sudden obstacles.

If an issue with the sensors or systems is detected, the in-built safety algorithm will ensure the bus comes to a safe and complete stop.

These capabilities have contributed to A2Z’s strong safety track record in South Korea, clocking over 668,000km of safe autonomous journeys from 2020 to May 2025.

Through this pilot, Grab also aims to explore how Singaporeans including platform workers can upskill and transition into emerging roles within the autonomous vehicle sector, while studying the potential higher-value career pathways that could arise as autonomous technologies evolve.

Grab plans to train up to ten safety drivers by the end of the year.

Five drivers including one GrabTaxi driver-partner, two GrabCar driver-partners, and two drivers from the National Private Hire Vehicles Association are currently undergoing training.

The autonomous shuttle will operate for two hours daily during off-peak periods, with additional hours dedicated to further training.

Grab and A2Z will closely monitor ride performance and safety outcomes before opening up more time slots.

Grab employees can track the shuttle schedule, vehicle location and, in the future, seat availability directly from the Grab app.

The onboard KartaDashcam devices will detect onboard passenger numbers and capture live images of vehicle location.

This data will be automatically synced with the Grab app, providing employees with the latest information.

Earlier this year, Grab signed a regional partnership with leading autonomous technology companies, including A2Z, Motional, WeRide, and Zelos.

Grab said the firm looks forward to participating in more autonomous pilots to understand the operational conditions for different autonomous vehicle services in Singapore, and be part of the nation’s efforts to improve transport connectivity through autonomous technologies.

Grab Malaysia, Proton expand strategic partnership to support e-hailing community