How high-def headlights will factor into autonomous vehicle systems
We talk about making vehicles smarter to make them safer. Why not go another step? What about making vehicle subsystems smarter to make driving safer still? What about, for example, smart headlights?
As autonomous driving technology advances, meeting safety standards and regulations continues to become more complex for automotive OEMs. At the same time, vehicle safety is a compelling market differentiator that can command greater margin. Given the trend that effective safety features will eventually be required on all cars, the sooner an OEM introduces safety features, the longer they can get value from them.
While autonomous cars will enable drivers to eventually become passengers, people will be in control of their vehicles for the foreseeable future. Advanced driver assistance systems (ADAS) are focused on improving vehicle safety by collecting environmental data, analyzing it, and acting on that analysis to improve the safety of not just the car and driver but also that of others sharing the road (i.e., other drivers, pedestrians, bicyclists, etc.)
Today’s vehicles collect environmental data through a growing array of sensors and cameras. Much work is being done to analyze this data to identify objects, track their motion, and predict their behavior (ADAS). It’s the last part — acting on information — that makes autonomous driving so challenging.
Ultimately, the ADAS will have the intelligence to use sensor and camera data to make the kinds of complex decisions required for autonomous driving. In the meantime, however, this same information can be used today to assist drivers in making safer choices.
One non-obvious innovation is to convey data to the driver via high-definition (HD) headlights. HD headlights employ a matrix LED. Individual control of LEDs enables the vehicle to dynamically adjust the headlights based on the driving situation and environment.
For example, glare to oncoming vehicles can be eliminated using beamforming and by segmenting light vertically (see Figure 1).
Figure 1: Glare to oncoming vehicles can be eliminated using beam forming and by segmenting light vertically. (Copyright Hella GmbH & Co KGaA)
The headlights can also be directed to follow the road (adaptive front-lighting system [AFS]) and highlight not just what is ahead but where the driver needs to see (see Figure 2).
Figure 2: Headlights can be directed to follow the road (adaptive front-lighting systems) and highlight not just what is ahead but where the driver needs to see. (Source: Cypress)
Rather than just illuminating the road ahead, smart headlights can also intelligently highlight objects in order to direct the driver’s attention to them. For example, when a road sign is identified, the headlights can not only eliminate the sign's glare, they can project an image of the sign on the road next to it — making it much easier for the driver to acknowledge it (see Figure 3).
Figure 3: Smart headlights can not only eliminate a sign’s glare, they can also project an image of the sign on the road next to it — making it much easier for the driver to acknowledge it. (Copyright Osram-Continental)
Similarly, the vehicle can use the predictive capabilities of the ADAS — capabilities that track the behavior of objects such as pedestrians and animals — to alert the driver of potential hazards (see Figure 4).
Figure 4: Smart vehicles can use the predictive capabilities of the ADAS — capabilities that track the behavior of objects such as pedestrians and animals — to alert the driver of potential hazards. (Copyright Osram)
These are just a few of the driver safety use cases. The same technology can also improve safety by alerting other drivers and pedestrians. For example, the rear taillights can project lights when backing up (see Figure 5). The ADAS can also alert other drivers that, based on their current speed, they are following too closely (see Figure 6).
Figure 5: Smart rear taillights can project lights when backing up. (Copyright Varroc Lighting Systems)
Figure 6: ADAS can alert other drivers that they are following too closely based on their current speed. (Copyright Varroc Lighting Systems)
A graphics MCU for my headlights?!?
With an LED matrix, a headlight can have more than a million pixels to control. In this scenario, the headlights use the road as a display surface — a surface which, in addition to the angle of projection, creates distortion. Being that the projection is effectively a human-machine interface (HMI), readability and clarity are a top priority.
In general, resolutions above 1K require a graphics controller. LED matrix designs need a graphics controller that can perform functions such as rotation, scaling, shearing, and de-warping (see Figure 7). As well, one that has a drawing engine (i.e., for arrows) and the ability to decompress stored images (i.e., road signs).
Figure 7: LED matrix designs need a graphics controller that can perform functions such as rotation, scaling, shearing, and de-warping. (Source: Cypress)
In addition to a graphics controller, HD headlights require processing capabilities, memory, the ability to generate multiple pulse width modulations (PWM) to drive LEDs, and the ability to interface to the ADAS and multiple light sources (see Figure 8).
Devices that integrate all of these capabilities together are known as graphics MCUs and coordinate operation of the headlights with the instrument cluster and ADAS. Controlling a vehicle’s light sources from one MCU enables complete coordination of the various functions. Thus, a road sign won’t be projected over a pedestrian’s projected path, rendering both unrecognizable.
Even more innovative and effective ways of alerting the driver become possible as well. For example, audible warnings could be coordinated with projected images to prioritize a particular hazard. Similarly, haptic feedback — such as vibration of the steering wheel or seat — could be used to trigger a more immediate corrective response from the driver.
To meet automotive standards, graphics MCUs need to have a higher ambient temperature rating (Ta = 125°C) than is required for subsystems like the instrument cluster (Ta = 105°C). Memory capacity and memory bandwidth need to be able to provide fast responsiveness so that the driver receives critical information as quickly as possible. Security of the system is important as well — given that the headlights are an essential element of effective automotive safety. Finally, these systems will need to be flexible. A sports car, for example, has a much different profile and responsiveness than a minivan or semitruck. With a programmable architecture, a single control system could accommodate a line of vehicles. Additionally, if a vehicle’s profile changes — for example, its wheels are raised — the headlights could recalibrate to adjust for the added height.
Figure 8: LED matrix designs need a graphics controller that can perform functions such as rotation, scaling, shearing, and de-warping. (Source: Cypress)
Graphics MCUs will help simplify the introduction of HD headlights and other non-obvious innovations into vehicular architectures. In many ways, HD headlights are the start of leveraging augmented reality technology for vehicle safety. Consider the level of driver distraction when listening to a navigation system or looking down at a cellphone for directions. Instead, the headlight controller could connect to the navigation system to non-intrusively project driving directions — as visual cues — on the road.
With high-definition headlights, automotive OEMs can substantially increase driver awareness of environmental hazards. At the same time, they can increase the visibility of a vehicle and its intentions to other drivers. All of these features can be delivered to drivers intuitively and seamlessly.
The end result is a safer and more efficient drive. And that’s a win for everyone.
Mathias Sedner works in product definition and marketing for Automotive MCUs at Cypress in Germany. He collects and consolidates customer requirements in order to define Cypress' next generation MCU products in the field of instrument cluster, HuD and lighting.
>> This article was originally published on our sister site, EE Times:"High-Definition Headlights: Next-Gen Automotive Safety."