Bringing safety to ADAS features and autonomous driving
How expertise in high-precision industries powers automotive safety
With the rise of advanced driver-assistance systems (ADAS) and autonomous driving in mainstream vehicles, the criteria for positioning and navigation technology are becoming more stringent—for both accuracy and safety. We’ll explain how combining measurements from satellite positioning and sensors can enable the precise positioning necessary for safe ADAS and autonomous driving.
Most vehicles come with satellite navigation as a convenience feature, which keeps you on track to your destination. This requires the car to know what road it is on and which way it’s heading—a level of accuracy measured in metres. This level of accuracy is not considered “high precision” compared to industrial applications of global navigation satellite systems (GNSS), like mining, precision agriculture, and survey, which measure accuracy in decimetres to centimetres. GNSS provides an absolute position, which determines your exact location in the world.
Also, common ADAS safety features, such as blind spot and lane departure warnings and adaptive cruise control, primarily rely on relative positioning sensors like RADAR, LiDAR and cameras, which do not provide absolute positioning (Figure 1).
Common ADAS safety features, such as blind spot and lane departure warnings, primarily rely on relative positioning sensors like RADAR, LiDAR, and cameras (Figure 1). These sensors are helpful for sensing objects around you, but do not provide absolute positioning like GNSS.
As the automotive industry moves to higher levels of driving automation (Figure 2), lane-level positioning that leverages both relative positioning sensors and absolute positioning from satellite signals is needed.
Level 0
No Automation
Driver performs all tasks related to driving.
Level 1
Driver Assistance
Driver performs driving tasks with assistance from built-in safety features. The driver maintains control of the vehicle.
Level 2
Partial Automation
Driver must remain alert, engaged in the environment and in control. Vehicle can perform one or more tasks simultaneously.
Level 3
Conditional Automation
The vehicle monitors the environment and can perform driving tasks. Driver is not required to remain alert but must be ready to take control.
Level 4
High Automation
Vehicle performs all driving tasks and monitors environment under limited conditions. Driver may still be required to intervene.
Level 5
Full Automation
Vehicle performs all driving functions under all conditions. Driver not required.
What are the challenges to safe ADAS features and autonomous driving?
The use case of enabling safe, precise positioning for automotive ADAS and autonomous driving versus other industrial applications presents several unique challenges, including changing environments.
Unlike most off-road or aviation applications, the on-road environment is highly dynamic (Figure 3). This includes changes in infrastructure, like tunnels, buildings, and overpasses, as well as overhead foliage. These obstructions impact GNSS positioning because the satellite signals are blocked from reaching the vehicle, affecting the ability to provide continuously available positioning information.
At the other end of the spectrum, roadways in more rural areas have very few objects and lane markings, rendering the information from relative positioning sensors, such as cameras, RADAR, and LiDAR, less useful. This is also true for challenging weather conditions, which can partially or completely blind the sensors with snow, rain mist, or fog.
Beyond the technical challenges, high-volume mass-market vehicle production requires positioning technology that is easy to integrate, globally available, and cost-sensitive.
How to bring safe, precise positioning to ADAS and autonomous driving?
To enable higher levels of driving automation, the vehicle must have continuous access to precise lane-level positioning, which means the information used to determine its position must be accurate, available, and reliable.
Accurate
Degree to which the position provided by a solution matches the true position.
Available
Percentage of the time that a solution can be used for positioning.
Reliable
The culmination of a positioning solution’s availability at the required accuracy level.
To achieve precise lane-level positioning, a combination of technologies is required. Over three decades, the engineers at Hexagon have applied their expertise in high-precision positioning to enhance accuracy at every component in the solution, including hardware, corrections, and software (Figure 4).
Regarding hardware, upgrading from traditional single-frequency and constellation to multi-constellation, multi-frequency capable hardware (antennas and receivers) to leverage the most GNSS signals is the first step. Adding in satellite and atmospheric error corrections, like TerraStar-X brings even greater precision. Finally, combining GNSS with information from inertial, visual, and vehicle sensors through sensor fusion and precise positioning engine algorithms improves not only accuracy but also availability and reliability through challenging conditions and dynamic environments.
For this reason, a combination of absolute positioning from GNSS and relative positioning from additional sensors is needed to provide continuously available high-precision positioning that enables the safety necessary for ADAS and autonomous driving. Read our eBook to learn the foundations of combining absolute positioning and relative positioning for safe automated and autonomous driving.
Looking to the future of ADAS and autonomous driving
Achieving precise positioning for automotive autonomy not only relies on leveraging knowledge and experience from the past, but it also needs to be flexible enough to incorporate future advancements. Building a foundation that powers safe ADAS features and autonomous driving through a reliable and available system sets your solution up for success. No matter what future innovations may come, Hexagon’s Safety Critical Systems engineers are up for the challenge to realise the future of autonomy for automotive.