HD MAPS How HD maps are changing the future of autonomous driving
From established GPS manufacturers like TomTom to newer companies like Waymo, High Definition (HD) maps are set to transform the future of autonomous driving. The HD maps market for autonomous vehicles (AVs) is estimated to be worth $1.3 billion in 2020, growing to $20.4 billion by 2030.
As self-driving cars continue to advance, so will the need for localisation technologies that can make decisions without human involvement.
HD maps help AVs to identify their exact position on a stretch of road and foresee the conditions up ahead, including any obstacles. They’re able to understand real-time changes in road conditions, anticipate traffic congestion and navigate accordingly to ensure the quickest route.
Let’s take a closer look at the role of HD maps in the future of autonomous driving:
How HD maps work
HD maps are designed in one of two ways. Either they use pre-existing road data, or the data is input by volunteers. Once the system is able to process the data, it can navigate autonomous vehicles to a specified location, anticipating traffic conditions and reacting to obstacles, much in the same way a human driver would in a normal car.
Through artificial intelligence (AI), the maps will also be able to update their own database in real-time as it gathers more data on road layouts. Through its connection with other devices, using Internet of Things (IoT), HD maps will be able to keep passengers up to date on traffic, weather and other obstructions that might delay a journey.
HD maps are configured from a number of sources, including sensors, which are mounted on the roof of the vehicle. The sensors deliver 700,000 data points per second and can pinpoint the exact location of objects up ahead. A 360-degree camera and GPS antennas monitor the performance of the vehicle’s components like the accelerometer and traction control.
The impact of HD maps on autonomous driving
Although autonomous vehicles won’t rely on HD maps alone, they will enhance the functionality of other technologies. For example, they can help increase the reaction speed of Advanced Driver Assistance Systems (ADAS) that are already in place in a vehicle. They will enable self-driving cars to make instant driving decisions, much quicker than a human can.
HD maps will also help to improve consumer trust in autonomous driving. According to a recent survey from the American Automobile Association (AAA), only 12% of drivers in the US would trust riding in an autonomous vehicle.
One area of concern for consumers is sensor failure. HD maps, however, will improve sensor perception in harsh weather conditions, recognising objects that may not otherwise be picked up by onboard sensors.
Aside from the safety aspect of HD maps, they will also improve the driver experience, making the journey more comfortable as alternative routes are taken to avoid delays.
The challenges of integrating HD maps into autonomous vehicles
One of the biggest challenges of integrating HD maps into self-driving cars is cost. Sourcing up-to-date data on road layouts takes time and increases production costs significantly.
A huge amount of data is also required to enable machine learning to understand complex driving challenges. The size of map datasets will continue to increase exponentially and require higher accuracy. In turn, this will have an impact on how software-over-the-air updates are received in terms of speed, bandwith and cost.
A significant amount of manual verification is also required to ensure that the data being collected is accurate. Managing quality control presents a challenge as maps need to be error-free.
There’s also an increasing amount of ‘red tape’ that has to be navigated before HD mapping technology can be approved. Before driverless cars are allowed on public roads, they’ll need to pass thorough testing from regulatory bodies.
The need for collaboration
HD maps require several components to operate. They also need to closely align with other products to ensure accurate data capture and system integration. To this end, the need for collaboration between automakers, component manufacturers and technology companies will only increase in the future.
Not only does collaboration bring costs down, it advances the development of autonomous vehicles on the whole. For example, TomTom has established partnerships with engineering company Bosch, as well as technology companies like Nvidia and Qualcomm.
Likewise, Toyota recently announced a collaboration with 3D mapping company Carmera with the aim of creating a proof of concept for a camera-based an automated camera-based HD map to be used on urban roads.
The more collaborations we see in the development of HD maps, the higher quality products will be and the easier they’ll be to integrate. Continued innovation will help realise the ultimate goal of level 5 autonomous driving.