NVIDIA Will Work With DeepMap To Improve Mapping for Autonomous Vehicles
Jane Lu
Oct 14, 2021 · 3 min read
As car technology develops, self-driving cars might be on the road in just a few years. Autonomous vehicles (AVs) would no doubt be useful as public transit vehicles or for transporting goods quickly across the country. There are still some aspects of AV driving that need to be addressed, such as special insurance considerations and vehicle mapping.
Startup companies focused on AV mapping solutions have been rising up quickly to help improve AV navigation. Green Car Congress reports that chipmaker NVIDIA has acquired one of these companies called DeepMap. How will this partnership improve AV mapping technology?
Self-driving cars are becoming better at navigating roads | Twenty20
NVIDIA and DeepMap collaborate on mapping technology
DeepMap already has a strong reputation in the AV mapping industry. It was founded five years ago by mapping technology specialists who were former employees at companies like Google and Apple. DeepMap uses AI to craft accurate and detailed AV maps.
The maps are created with car-mounted cameras, which transmit sensor and radar data to the car's computer. That data is stored in the car's memory bank so that it can automatically make steering adjustments as needed.
DeepMap’s software will help enhance NVIDIA DRIVE—a vehicle update program. The goal of this acquisition is to continually enhance the localization of AV mapping technology for new roads and closures. This will help AVs know their precise location and where they’re going.
Can autonomous vehicles really follow maps without a driver?
GPS mapping has been around for years, and some cars even have their own built-in navigation system. We also know from semi-autonomous driving technology, like adaptive cruise control, that vehicles are capable of steering themselves. However, these programs don't run with road closures or similar conditions in mind.
AVs also lack the intuition of a human driver. For example, when driving through heavy rain, a human driver might slow down because the roads will be slick. In fact, bad weather can make it hard for AVs to navigate; rain or fog affects their sensors and ability to "see" the road.
How can AV mapping be improved?
Mapping technology needs to be precise for consumers to feel confident riding in a self-driving car. One way to accomplish this is by implementing constant updates for the car's mapping system. If you've ever used an outdated GPS, you know how frustrating it can be to be steered down a closed route.
AV mapping has to function seamlessly no matter where you are in the world, and be accessible even in areas without internet access. The maps will have to account for work zones and lane closures.
You also don’t want to worry that your car might drive you off a cliff or into a potential hazard. This is especially a concern for companies that want to use AV technology on commercial fleet vehicles. Companies will want to know that their goods will arrive safely and on time.
NVIDIA and DeepMap's collaboration will help enhance the current state of AV mapping technology. Once AVs are actually released, they will need special insurance for accidents or mapping malfunctions.
In the meantime, you can use Jerry to save money on car insurance. The free app compares the most up-to-date car insurance rates so you’re always getting the best coverage for the best price.
