One of the most interesting and fastest growing safety applications is Advanced Driver Assistance Systems (ADAS). Learn more about ADAS and design concerns related to two of its most important features: back up safety cameras and around-view.
The importance of automobile safety has been the force behind a number of key innovations over the years from anti-lock brakes to airbags. German auto-makers have been first to leverage many of the most advanced technologies to improve driver safety and the overall driving experience. For example, over the last two decades alone, German vehicles were the first to offer active lane keeping assist, which introduces a vibration to the steering wheel when the driver veers from his or her lane. Then came side assist, a radar-based system that alerts the driver to any vehicle in the blind spot, helping her change lanes safely. This was followed by the introduction of stop-and-go and brake assist, an extension of cruise control, which reduces the risk of rear-end collisions by quickly asserting the brakes before an impending crash.
These features have become more sophisticated over time and more pervasive, particularly with the introduction of safety cameras in the vehicle to literally eliminate the driver’s blind spot.
The Insurance Institute of Highway Safety in the United States recently released a study that showed the chances of dying in a crash in a late-model vehicle have fallen by more than a third in the short span of three years. In fact, advances in safety have improved fatality rates by such a high margin, that cars being sold today offer meaningfully better safety statistics than even those introduced only five years ago. This is creating a virtuous cycle for car makers, giving drivers a reason to purchase a new vehicle sooner rather than later.
There are many reasons for the improved safety, from mechanical and structural improvements to increased safety-related legislation. But one of the most interesting and fastest growing safety applications is Advanced Driver Assistance Systems (ADAS). These systems use a combination of sensors, cameras and displays to provide greater driver visibility and also react to dangerous situations when a driver doesn’t. A blind spot is defined as the area where a driver’s view is obstructed. The back-up camera, mounted at the rear of the car with the display mounted in the rearview mirror or dashboard, has become a staple in mid-to-high-end vehicles to provide driver visibility behind the vehicle and reduce accidents. The National Highway Traffic Safety Administration (NHTSA) estimates that rearview cameras can reduce back-over fatalities by over 50%. This feature became a mandated safety requirement in the U.S. when government regulation was passed requiring back-up cameras for all new vehicles beginning in 2018.
Similar regulations are being considered in Europe and Asia, and car makers are already adopting rearview cameras across models. The TW9992 is a good example of an automotive video decoder IC addressing this global safety trend. It takes single-ended and differential composite video inputs from a car’s rearview camera and applies its automatic contrast adjustment feature to reduce or boost image brightness/contrast for greater visibility and safety. The decoder then outputs the video over its MIPI-CSI2 interface to the head unit’s SOC, which processes the video and sends it to the dashboard’s LCD screen or rearview mirror.
With rearview and other cameras becoming mainstream, cost and reliability are the two critical design concerns facing auto-makers. The two elements to the system cost equation are the camera and the display. Initial implementations leveraged a low cost display in the rearview mirror. As navigation units and in-dash displays have become more common, they are being used to display the rearview camera video, as well. The original designs attempted to run this video through the core processor managing the infotainment subsystem. These sophisticated SoCs introduced reliability issues, booting up too slowly to provide timely rear camera video to the driver immediately after initial ignition of the vehicle.
The alternative, particularly as the load on these SoCs continues to increase, is to leverage what looks like a “hardwired” video decoder and LCD controller solution that bypasses the SoC to deliver near immediate back-up video to the display, ensuring reliable delivery in less than 500 milliseconds.
This approach allows the system to detect a lock-up and bypass the SoC when necessary. The hardware bypass provides the communication integrity and reliability demanded by every major auto OEM.
An additional challenge faced by today’s system architects is that most digital SoCs don’t have enough ports available to accept the growing number of video inputs that modern vehicles require. So, while they have the horsepower to process this data, albeit not always as reliably, there is no way to immediately get the data from multiple inputs into the device. The answer is interleaving the signals together over a single data bus that requires only a single video interface to the application processor. The SoC is then able to demux and process multiple different video streams. Over time, the number of interleaved streams will increase, enabling more video inputs. The multi-channel analog video decoder with interleaved digital output approach frees up valuable pins on the SoC, which is a key design consideration.
There are also cost advantages. With as many as eight cameras per car, cost is becoming a real factor as these safety features are introduced downstream into more mid-and low-end vehicles. High definition cameras requiring more complex and expensive processing and cabling is not cost-effective for these vehicles. Whereas, analog technology is full featured and provides reliability advantages at a much lower implementation cost.
With auto-makers no longer debating the usefulness of rearview cameras, most are making significant advances to enable around-view capability as a standard safety feature across car models.
An around-view monitor system, also referred to as a surround-view monitor, processes video from four cameras and then combines the four images into a single birds-eye, top-down view as if a camera was stationed directly above the vehicle. The monitor helps drivers visually confirm the vehicle’s position relative to surrounding objects for easy maneuvering and parking. According to Infinity Research, the around-view monitor application alone is growing a compounded 33% through 2018.
This application has become more cost effective to implement and provides a number of reliability benefits through new multi-channel video decoder technology specifically developed for around-view parking assist applications. Integration has been a key enabler, with the latest solutions integrating four high quality NTSC/PAL analog video decoders with 10-bit analog-to-digital converters (ADCs) to support four independent analog camera inputs simultaneously. A flexible digital output interface sends the images to a processor that can then combine the four images into a single unified around-view image. The on-chip analog video encoder provides the ability to transmit the combined video as a standard analog composite signal to the head unit display. The ability to directly receive differential analog video inputs eliminates the need for an external op amp on each input channel, further reducing component count. This high level of integration helps to simplify the system design and minimize the solution footprint to preserve critical board space and reduce the overall system cost, a key factor in enabling this capability to become more pervasive.
Flexibility is also important. Given the various implementations possible, these new four channel video decoders can now offer a digital output interface to support multiple configurations, including a standard ITU-R BT.656 format output on four 8-bit buses at 27MHz; time multiplexed byte interleaved output to send two channels on a single 8-bit bus at 54MHz; or all four channels on a single 8-bit bus at 108MHz. This gives the system designer the ability to create a number of variations for different models without a completely new design. Designers should also be wary of video decoders and encoders with features that are more suited to building surveillance than an automotive application as they are missing key diagnostic features such as short-to-battery and short-to-ground, that further improve reliability and ensure system designers aren’t paying for features they don’t need.
Safety has become a significant differentiator over the last decade – the availability of crash test data and the ability to avoid minor dents and dings as well as serious accidents has become a real factor in consumers’ choice of vehicle. This has changed how the auto-makers are marketing these new features. With consumers prepared to pay for safety, features like blind spot monitors and lane departure warnings are proving to be more compelling to justify the presence of a central LCD and infotainment system than paying for the upgrade to get a better radio or navigation system. Premium additions to basic car models are increasingly centered on enabling better visibility and safety, and are proving to be a better economic proposition for the car maker as well.
There’s more on the horizon as the cost of implementing safety features makes it possible to deliver them in more mainstream models. More and more video will be used to deliver capabilities like night vision and auto-brightening and darkening to improve visibility, for example, and the windshield will become yet another display for diagnostic data. The ability to deliver reliable, easy to integrate and cost effective video solutions will be key enablers of the next generation of ADAS features, which are increasing driver safety and ultimately consumer demand for new vehicles.