New high-accuracy GPS aims to end navigation woes

The familiar experience of watching your Grab driver’s vehicle icon wander aimlessly on the map, especially in Singapore’s central business district (CBD) and underground portions of expressways, may soon be a thing of the past.

Ride-hailing giant Grab has launched a pilot programme in the city-state to bring lane-level GPS accuracy to its drivers in a bid to address inaccurate location-pinning issues that have long plagued navigation in dense urban areas.

The initiative, rolled out together with smartphone maker Oppo, chip designer Qualcomm Technologies, and positioning tech firm Swift Navigation, marks the first deployment of such high-accuracy GPS on mobile phones for a ride-hailing app in Southeast Asia.

In areas with high-rise buildings, multi-level roads and underground tunnels, standard GPS accuracy can degrade by over 20 meters. As such, drivers may miss turns or arrive at the wrong pick-up point, resulting in inaccurate estimated arrival times (ETAs), and ultimately, cancelled rides.

The new system promises to reduce that margin of error significantly, allowing drivers to pinpoint their location with far greater precision.

“Driver- and delivery-partners will be driving with technology that is at the cutting edge – similar to high-precision positioning technology used in autonomous vehicles and even motorsports,” said Nilofer Christensen, head of consumer product at GrabMaps.

The pilot, which started in October 2025, involves nearly 250 select driver- and delivery-partners. Around 60 of them have been equipped with Oppo Find N5 foldable phones, which feature the necessary hardware to support the new technology.

The system works by combining several connectivity and software technologies. The Oppo phone’s dual-frequency GPS capabilities capture a wider range of satellite signals, which are processed by Qualcomm’s Snapdragon 8 Elite mobile platform that receives real-time correction data.

The correction data is supplied by Swift Navigation’s cloud-based Skylark service, which refines GPS signals by cross-referencing them with a network of ground stations, including the Singapore Land Authority’s satellite positioning network.

When driver-partners access the Grab driver app and use GrabMaps for navigation, the app receives high accuracy location, improving their navigation experience in real-world.

Grab driver-partner Santho Rudy Peter, who is taking part in the trial, said: “In the past, GPS signals could get quite weak in the CBD, especially around tall buildings where the signal bounces around. With this pilot, the location accuracy feels much sharper and more stable, and navigation updates happen almost instantaneously. It really helps me get to pick-up points faster, with less guesswork.”

For consumers, the improvement translates to more reliable ETAs and a smoother experience, whether it’s a ride from the correct entrance of a large mall or a food delivery to the right condominium tower.

Apart from the overground pilot, Grab is also looking to solve the challenge of underground GPS blind spots. The company has been using its proprietary KartaCam and light detection and ranging (Lidar) technology to map basement car parks of hotels and shopping malls in the Marina Bay area, such as the Ritz-Carlton and Millenia Walk.

This allows for a seamless handoff from the high-accuracy GPS above ground to a dead-reckoning system used in underground navigation when the vehicle enters a car park, preserving turn-by-turn navigation.

While the enhanced positioning capability in a pilot phase, Grab plans to eventually extend this enhanced positioning capability to its own Karta devices to benefit more partners.

“This pilot is the first step toward Grab's vision of lane-level navigation in Singapore and beyond,” the company said in its statement. “Grab is laying the foundation for smoother, faster, and more reliable journeys across Southeast Asia.”