■ Driver Assistance
lane-change assistant — to name just one ex- ample — can help to compensate for a dri- ver’s limited mobility and restricted vision,” says Zittlau. Using radar sensors to monitor traffic to the right and left of a car, this system warns the driver of any objects that pose a danger should he or she wish to change lanes or execute a right or left turn. In addition, be- fore the doors are opened to leave the vehi- cle, the same sensors also warn of approach-
where our stop-and-go assistant comes in,” explains Michael Lütz, head of Project and Product Management at Driver Assistance Systems. “It automatically accelerates the car or slows it down to a complete halt if neces- sary. Using a CMOS video camera, the system constantly monitors the vehicle in front of it. A computer then coordinates the operation of the vehicle’s brakes and engine.” If the vehicle in front changes lanes, the camera automati-
ing cyclists, skaters or cars concealed in the mirror’s blind spot.
cally focuses on the next vehicle ahead and resets the optimum distance and speed.
Another factor spurring growth in this field is the growing need for enhanced driving comfort. Anyone who has ever spent a long period in congested highway traffic knows the stress of having to switch constantly be- tween the gas and brake pedals. “That’s
The Lane Departure Warning (LDW) sys- tem functions according to the same princi- ple. Here, a video camera monitors lane markings. If the software detects that the ve- hicle is in danger of crossing the white line, it produces a clearly discernible vibration in the
Variable cruise control, a traffic-jam assistant and LDW help drivers, who may also deactivate the systems at any time.
steering wheel so as to warn the driver. In the future, the LDW could also be transformed into a camera-controlled lane-guidance sys- tem, which would automatically correct the steering whenever the vehicle threatens to stray off course.
In another highly useful development, Siemens VDO has come up with software that
FOLLOW THAT CAR! — NEW DRIVER ASSISTANCE SYSTEMS FROM SIEMENS’ LABS
In combination, three sensor systems calculate the probable position of the vehicle ahead
It’s stop-and-go time on the highway again, and all that braking and ac- celerating is really stressing drivers out. To help prevent rear-end collisions in congested traffic, Siemens is developing a system that can maintain a con- stant distance from the vehicle in front, automatically braking and accelerat- ing the car as required. What’s more, the system can even do this when driv- ing around corners, as demonstrated by a test vehicle at the end of April 2005 in the “Invent” project sponsored by the German Ministry of Education and Research to promote user-friendly technology for intelligent transporta- tion. The car is fitted with a video camera and an infrared laser (lidar), mounted on the windshield, plus radar sensors fitted behind the front
independently accelerate or brake the vehicle. fender. “The various sensor systems and the areas they scan all complement one another,” explains Dr. Georg von Wichert from the Competence Center for Autonomous Systems at Siemens Corporate Technology. As Wichert ex- plains, this mix of different sensors increases the power and reliability of the system as a whole: “After all, there’ll always be times when individual sen- sors are impaired by ambient conditions.” The image-processing system hooked up to the video camera is able to calculate — on the basis of their distance ahead — whether vehicles in front are accelerating or decelerating. In addition, the system also registers road markings — which serves to de- termine, among other things, lane width and curvature. Using this informa- tion, a motor fitted to the steering column of the test vehicle is able to hold it automatically in the middle of the lane and even follow the car in front as it drives around a corner. In combination with a lidar sensor, a radar system measures the distance to, and the speed of, vehicles in front. Unlike today’s systems such as automatic cruise control (ACC), which work best at high speeds and are unable to steer the vehicle, this new system can control the car fully automatically at low speeds of between zero and 50 kilometers per hour. This is made possible by special software that checks the data from all three sensors — video, radar and lidar — millisecond intervals in order to verify that their readings do not contradict a model of the vehicle environ- ment. The latter features descriptions of probable vehicle movements, as based on the laws of nature. If, for example, the sensors have registered a vehicle in front that is driving at a certain speed, the model tells the system that the vehicle will probably continue to travel at that speed. Because the system knows its own speed, it can calculate where the vehicle in front should be by the time of the next reading, which in turn is compared with the new sensor data. In this way, the system can determine the precise position of the vehicle in front and therefore ensure that it remains at a safe d i s t a n c e . O f c o u r s e , t h e d r i v e r c a n a l s o o v e r r u l e t h e s y s t e m a t a n y t i m e a n d ■ R o l f S t e r b a k
ROBOTS THAT CAN RECOGNIZE LOCATIONS
Researchers at Siemens Corporate Technology in Munich are busy looking at how machines can develop a sense of direction in unfamiliar surroundings. For some time now, Michael Fiegert and his team from the Competence Center for Intelligent Autonomous Systems have been work- ing on self-propelled robots for applications such as transport systems in hospitals and cleaning equipment in supermarkets and warehouses. “The same principles can be applied to other mov- ing objects that utilize visual-recognition systems. In the future, this might include cars equipped with cameras,” says Fiegert. At present, Fiegert’s team is using laser scanners to generate 3D im- ages of a robot’s immediate surroundings. In the future, however, these will be replaced by cheaper video cameras. The principle is the same. Images of certain places are taken from different perspectives and at varying resolutions. On this basis, characteristic features are determined and then saved in the form of a code. The same characteristic features are then coded from another angle in order to calculate their 3D posi- tions in the space under consideration. Using this method, a mobile machine is able to work out its position in relation to its environment. Whenever it returns to a given location, the machine recognizes it on the basis of the characteristic features saved in its memory. Siemens has recently installed this technology in a driverless forklift designed to independently transport pallets around a w a r e h o u s e . T h e p r o j e c t i s a l r e a d y s o a d v a n c e d t h a t F i e g e r t i s c o n f i d e n t t h a t t h e s y s t e m w i l l b e ■ R o l f S t e r b a k used in practice in the near future.
Virtual road signs. In the future, impor- tant traffic information could be projected on the windshield.
can read speed limits on traffic signs. If the driver violates a speed limit, the system dis- plays a visual warning signal. “But it won’t au- tomatically reduce your speed,” Lutz empha- sizes. “Furthermore, the driver can always
deactivate the system,” he adds. In addition, there are now plans to extend the system to cover stop and yield signs. Meanwhile, an- other application from Siemens VDO — ad- justable cruise control (ACC) featuring vari- able speeds — utilizes a new light-based detection technology to measure distance. This so-called lidar sensor is comparable to
conventional radar sensors, but is substan- tially cheaper. Moreover, unlike radar sensors, which have to be fitted behind the radiator grille or the fender in order to protect them from the elements, lidars can simply be in- stalled behind the windshield.
Even at night, Siemens VDO helps keep drivers fully in the picture with two night-vi-
AUGMENTED REALITY ENHANCES NAVIGATION SYSTEMS
Navigation system calculates route and current position
Onboard camera captures live picture of the view in front of the car
3D map knows topography of area
AR system calculates 3D view of the recommended route
Navigation system display shows combined (AR) camera view and 3D route
Navigation systems featuring augmented reality could soon be provid- ing drivers with enhanced assistance. With this system, the driver’s actual view of the road ahead is supplemented by virtual, computer-generated in- formation. “We’re currently working on a prototype that may well develop into a product in a few years,” says Dieter Kolb, a software engineer at Siemens Corporate Technology. Today’s systems present route information as a pixilated graphic or a map representation on the display unit, accompa- nied by voice output. In the future, however, the route could be displayed as a colored strip superimposed on images of the road ahead, as captured by an onboard camera. The driver would then see the route as if it were literally painted on the road. And because the display would show the same view of the road ahead that the driver sees through the windshield, motorists would have no trouble orienting themselves. Moreover, the “pathfinder” provides
systems but also increase road safety. clear routing information at night or when the driver’s view of the road ahead is obscured by other vehicles. For example, the system shows the driver the course of the road after a hill or a tunnel and therefore gives advance warning of any sharp bends or other potential trouble spots. In the next project stage, route information will be projected directly onto the windshield — although only when required, so as to guarantee a free view of the road at all other times. If navigation assistance is needed, the driver simply presses a button on the steering wheel in order to project routing information onto the head-up display on the windshield. A major advantage of augmented reality is that the driver always keeps an eye on the road ahead even when consulting navigation data. In other words, this solution w i l l n o t o n l y i m p r o v e t h e e r g o n o m i c c h a r a c t e r i s t i c s o f t o d a y ’ s n a v i g a t i o n ■ R o l f S t e r b a k
Pictures of the Future | Fall 2005