Research studies show that six out of ten front-end crashes would not occur if the driver could react a split second earlier. Matthew Beecham reports on some technical advances in driver assistance systems that could help cut the accident toll.
It makes grim reading but, on average, someone dies every minute somewhere in the world as a result of a traffic accident. The European Commission estimates that the 1.3 million road accidents per year in Europe cause about 1.7 million injuries and some 40,000 fatalities at an estimated cost of €160 billion. The EC has therefore established its so-called e-Safety programme which aims to halve the number of deaths from traffic accidents by 2010.
“The EC target to reduce the road mortality rate by half by the end of the decade is certainly an ambitious goal,” said Dr Rainer Kallenbach, executive vice president for Bosch’s automotive electronics division. “Driver assistance systems, however, are still in their infancy with not many cars equipped with it. Therefore they can only make a contribution toward meeting this goal. However, we see that increasing fitment of ESP [electronic stability program] will make an even bigger contribution to cutting the number of fatalities on European roads.”
Bosch pursues the aim of reducing the frequency and severity of accidents by developing active and passive driving assistance systems. The company sees its driver assistance technology filling the gap between safety and comfort. “We are approaching it from the comfort side, adding features to assist the driver. The driver maintains full responsibility for what he is doing,” said Kallenbach.
According to Bosch, driver assistance systems aim to make the vehicle capable of perceiving its surroundings, interpret them, identify critical situations, and assist the driver in performing driving maneuvers. The object is, at best, to prevent accidents completely and, at worst, to minimise the consequences of an accident for those concerned.
Last year, Bosch introduced its first generation Predictive Safety System, which uses information from the adaptive cruise control (ACC) to prepare the safety systems in a car by pre-boosting brake pressure. Other systems demonstrated by the company recently include full speed range ACC (known as ACCplus and due to enter series production next year), lane departure warning systems and automatic and semi-automatic parallel parking systems.
To help parking maneuvers, Bosch’s semi-autonomous parking assistance system works by measuring the size of a parking space and then telling the driver what steering movements to make. In a more advanced version of Bosch’s system, called Park Steering Control, the steering maneuvers, once calculated, will be performed by an electronically controlled power steering unit. “As soon as the system is activated, the driver releases their hands from the steering wheel but continues to operate the accelerator and brake pedals as the car moves back into the parking space. As soon as the driver decides to steer, the system shuts off,” said Kallenbach.
“We are not talking about complete parking system automation because this would mean the driver relinquishing all braking and acceleration to the computer. While we can do such things today in research cars, that represents a huge step. It means that the driver is just watching and no longer taking active decisions. So we see the introduction of such technologies will be in smaller, evolutionary steps.” He added that Bosch’s Park Steering Control will be ready by 2007. “Although we can give dates for readiness for the start of production, the final decision rests with the vehicle manufacturer.
Hella is also supportive of the European self-ruling to reduce the number of deaths on EU roads by 50% by the year 2010. The German lighting and electronics specialist is busy engineering driver assistance systems so that it can move rapidly down from the initial high-line applications through to smaller and less expensive cars. “Hella has a strong commitment to support the European Commission’s e-Safety programme,” said Winfried Menge, Hella’s head of marketing, automotive electronics. “That means our target is to develop driver assistance systems not only for premium cars but find the right price/performance relationship in order to bring these types of technologies and products at least into the upper mid-class segments.”
Hella’s so-called Lane Change Assistant monitors adjacent lanes during overtaking and changing lanes. Drivers are warned of critical situations by optical, acoustic or haptic signals. The system works by using two 24 GHz radar sensors on both sides of the vehicle to recognise other road users or objects behind and alongside it. The sensors have a range of 50 metres and are not affected by weather conditions. If there is a vehicle in the adjacent lane and the driver intends to change lanes, he/she will be warned either by an optical signal (e.g. in the wing mirror) an acoustic signal (e.g. a warning sound in the radio loudspeaker) or a haptic signal (e.g. ‘shaking’ of the steering wheel).
Hella’s ACC uses infrared sensor technology. The system works by scanning vehicles in the same lane traveling in the same direction and then determines the distance between the vehicles and their relative speeds. By adjusting engine performance and braking force accordingly, the ACC system ensures that the distance to the vehicle in front remains constant in relation to the vehicle’s own current speed. The ACC system is based on infrared laser technology and is an extension of the conventional speed regulator.
The company points out that the performance, functional ability and reliability of infrared laser technology is comparable with the more expensive radar technology. The infrared proximity sensor used for ACC is based on opto-electronic measuring technology: the time it takes light to travel a certain distance is determined following the principle of light detection and ranging (LIDAR). For this, a brief light impulse is transmitted and the return scatter signal is recorded with the aid of evaluation electronics.
“We shall introduce LIDAR in early 2006 with a North American customer,” added Menge. “It will have the adaptive cruise control. We also have our 24 GHz radar lane change assist, which will go into production in late 2005 in Europe. Our third main technology is the rear view camera. This will go into production in 2006, initially for the US market. By July 2006, we will be in production with all three technologies. There is a fourth one. Start of production for our automatic parking assistance system has not yet been announced. We are still in discussions with the customers because together we want to identify the best price/performance ratio. We aim to establish from the OEM what a ‘must’ (with regard to functionality) is and then what should be the price range for the consumer.”
Hella decided to develop an infrared-based system over radar-based systems due to its long established strength in optical systems. “We found that we are able to provide a control performance for a LIDAR ACC sensor which is at least as good as a radar system, especially in poor weather,” said Menge. “In other words, to a certain extend it can see through fog, rain and spray. But in addition the system is able to assess the visibility and provide a warning in case of too limited vision. Above all, LIDAR sensors are about 50% the cost of a radar sensor. That means we have a viable solution for small- and mid-range cars, not just premium models.”
Continental Teves has also integrated a number of its driver assistance technologies into its so-called Active Passive Integration Approach (APIA), which links steering and brake safety systems with passive devices such as seat belts and air bags to minimise crash dangers. “We are coming at it from comfort systems toward the more related safety systems based on our braking system technology,” said Georg Otto Geduld, senior manager, Advanced Development Product Line, Driver Assistance Systems, Continental Teves. “We are integrating our braking technologies with sensor technology to provide driver assistance systems. APIA is a system approach. It represents a synergy of the company’s knowledge. Our APIA approach network’s all of this technology into a single system.”
As far as TRW’s stage of development with regard to its Driver Assistance System is concerned, the company decided that it needed a number of key enablers to allow it to enter the driver assistance market. “The first of those was our long range ACC radar sensor of 77 GHz. We are just about to launch the second generation with a customer,” said Ken Lowe, senior product manager, Driver Assistance Systems and Steering Europe, TRW Automotive. “It is a more compact and lower cost product than the initial AC10. Development work on our third generation technology has already begun. We are aiming to introduce it in 2009.”
TRW Autocruise’s new generation AC20 radar system is half the size and weight of TRW’s current AC10 model at a reduced cost. The AC20 radar system offers several enhanced driver assistance function options such as ‘follow to stop,’ ‘collision mitigation’ and will support future assisted ‘stop and go’ systems. The ACC is a conventional radar based adaptive system which acts as normal cruise control holding the vehicle at a set speed – until a slower vehicle appears in front, when it automatically accelerates and brakes the vehicle to keep a driver-selected gap (constant time interval) behind the slower vehicle. The information from the 77 GHz radar is analysed by electronics contained in the ACC unit. However, the new radar and enhanced transmit-receive module captures data for an additional 50 metres compared with the previous system – extending its range to 200 metres.
The new radar’s control unit also offers enhanced driver assistance features including follow to stop, where the system brings the vehicle to a complete standstill. Assisted stop and go, an optional feature planned for the future, will automatically stop and accelerate the vehicle in heavy stop and go traffic. An additional planned future option – emergency brake assist – will automatically apply the vehicle’s brakes to prevent any reaction delay by a driver suddenly confronted with an obstruction.
Cambridge Consultants Ltd (CCL) have also been busy pushing back the technical boundaries of driver assistance systems. CCL has combined a video link with their 5.8 GHz pulsed-radar anti-collision system to provide an enhanced product.
Today’s current video-based collision systems deliver generally distorted images, which are hard for the driver to read, but by adding radar technology they can offer highly accurate detection of object distance and trajectory in the form of an overlaid image on the screen in which the danger point is indicated. Further processing of the radar data with ‘object classification’ algorithms can provide highly intelligent specific alerts, such as distinguishing a pedestrian from a lamppost.
Peter Bell, commercial director, Automotive & Transport, CCL discussed what is in the pipeline in terms of anti-collision systems. He said: “CCL is currently working on both lateral and forward accident mitigation systems. Although we cannot mention names, we can say that we are actively working with major players in the automotive industry and expect to see these on cars in the near future.”
In terms of the enabling technologies, the addition of radar and laser detection to conventional cruise control is seen as the key to a whole army of new sensors capable of improving the driver’s knowledge of his immediate surroundings. Bell agrees that the two key technologies are radar and vision. “Radar gives a very precise measurement of position and speed. However, radar is not very good at distinguishing different types of objects. Therefore, second generation systems will need to fuse with vision. The pre-crash strategy can then take into account, whether the car is about to hit a street-lamp, versus a pedestrian versus another vehicle. I don’t believe any sensor on its own will provide the complete anti-collision system. Vision is not very good at giving precise 3D location or speed information. Mono-vision cannot do it at all, while stereo-vision struggles given the short baseline. Laser-based systems do offer an alternative to radar, however they are more sensitive to environmental problems. I think that long term they are less likely to succeed.”
While active driver intervention clearly holds far more possibilities, it is also fraught with difficulty. Can and should we rely on these systems? Bosch’s Kallenbach said: “This is certainly a big issue but there are no easy solutions. For the time being, we believe that the ultimate responsibility must stay with the driver. This is our strong philosophy. Yes, we want to support him but in the end he must take responsibility for his actions.”
TRW’s Lowe agrees that the driver relinquishing all control to an onboard computer is still some way off. He told just-auto: “It is technically feasible to do those things but there are other issues as to why it is not appropriate to do it right now. You will see it happening on the braking side first before you start to see automatic steering intervention.”
“Generally speaking, [driver assistance] technology shall never take over responsibility for driving,” added Continental Teves’ Otto Geduld. “The driver will always have full responsibility. Assistance implies supporting the driver during his tasks to make his own driving more safe and comfortable.”