1. Protection of vulnerable road users

At night, in bad weather conditions or on country roads, the risk of accident is especially augmented. It is thereby particularly important to protect vulnerable road users like pedestrians, bicyclists or motorcyclists.

By means of far infrared or near infrared cameras, we can detect vulnerable road users in front of the vehicle in time. The sensor data is then evaluated by means of image processing. In addition, a lidar is applied as a supporting sensor measuring the distance.

Within the automotive industry, first realizations of such a pedestrian detection can already be found, for instance in the application “Night Vision” by BMW or within the “Honda Intelligent Night Vision System“.
Furthermore, such functions could be used for the identification of persons within the area of civil security.

2. Vehicle recognition

An important precondition for being able to realize safety functions within a vehicle is to recognize other traffic participants like cars or trucks, as well as their position and driving direction.

Within our concept vehicle Carai, this happens by means of different sensors, which can either be applied individually or by sensor data fusion. Next to radar and lidar sensors, camera systems of all kinds are applied (near infrared camera, far infrared camera, 3D camera, stereo camera, gray scale value camera, color camera). The data is evaluated by means of image processing and signal processing, either on the basis of raw data or of already preprocessed measurement data. Further, tracking and classification algorithms are applied.

The function of vehicle recognition has already been realized in several applications, for instance in Adaptive Cruise Control (ACC) or the extended ACC function of Stop-and-Go, which automatically controls stoppage and restart in traffic jams. Likewise, the so called Collision Mitigation can be realized by means of vehicle recognition. Such a brake assistant system mitigating collisions can for instance be found within the „Collision Mitigation Brake System“ by Honda.

3. Obstacle recognition

In some traffic situations, it is sufficient to be able to detect moving or non-moving obstacles without classifying them. Rather, the focus is on mere detection of obstacles and their velocity vectors. By means of a corresponding driver assistant system, it can quickly be decided what size an obstacle is and whether for instance braking is necessary. Such a function of obstacle recognition can also be used for free space detection, for example in parking situations.

Within the concept vehicle Carai, this is realized by means of lidar and radar sensors as well as different camera systems. Next to video, infrared camera and 3D camera, we also apply ultrasound technology, which is quite economic for automobile manufacturers.

The function of obstacle recognition can be used for parking assistant systems or in detecting road work barriers. Another possible application are so called Pre-Crash systems, like they are already applied for instance by Daimler. In the future, a usage in the area of Collision Mitigation would also be imaginable.

4. Lane detection

Precondition for being able to gain an in-depth knowledge about the vehicle’s surroundings is certain information about the lane: Where is it located? What does it look like? How wide is the lane? Based on road markings or by means of other characteristics (e.g. the texture of the surface) information like these can be generated.

Within the concept vehicle Carai, this is put into practice with the aid of lidar sensors and video cameras, which detect road markings and the edges of the lane. This information is evaluated by means of tracking. This way, the future course of the lane can be estimated and possible changes like curves or a changing width of the lane can be predicted.

Practical applications can be found within systems like Lane Departure Warning or Lane Keeping. The automobile manufacturer Honda for instance already puts into practice a system called „Lane Keeping Assist System (LKAS)“. Adaptive Cruise Control (ACC) next generation or lane change assistants would be further applications using this function of lane detection.

5. Localization

By means of localization one can gain information about the vehicle’s position within the road. On a first level, it can be assessed on what kind of road the vehicle is moving, for example on an autobahn or a country road. Next, the vehicle can position itself laterally, meaning within the lane, as well as lenghtwise, in order to gain information for instance about the distance to the next exit. The aim is a high-precision location, either with or without a-priori knowledge (e.g. through digital maps).

The localization of the Carai can be realized by means of camera systems and lidar sensors. Furthermore, GPS/ DGPS is used for approximate positioning, while precise localization accurate to the centimeter can be obtained by recognizing land marks (e.g. road markings, traffic signs or traffic lights) in connection with digital maps.

Localization can be used in applications like a map supported Adaptive Cruise Control (ACC) or for Location Based Services, which display information for instance about traffic jams only for the vehicle’s current position.
The project “SAFESPOT” funded by the European Union is likewise working on the basis of such functions.

6. Communication

Exchanging information and data between different road users and the infrastructure is considerably contributing to improving one’s environment perception and thereby augmenting the road safety. This can happen either in the sense of car-to-car-communication or from a vehicle to any other road user, for example a pedestrian (car2x/ c2x respectively vehicle2x/ v2x).

In order to realize this function, active communication devices like GSM, standard communication devices (WLAN) or dedicated communication devices (e.g. based on actual C2C-standards, like 802.11p or IEEE 1609.3/4, WAVE) are applied within the concept vehicle Carai. These can be used individually or in fusion with other sensors. In terms of algorithms, a general exchange of information about certain communication protocolls is taking place, which can then be intergrated into one’s own situation perception.

Practical applications can be found within the area of pedestrian protection as for instance in the project “WATCH-OVER”. Likewise, applications like the so called eCall are realized by means of communication functions. Systems like the “BMW-Assisst” are able to automatically make an emergency call if needed.

7. Interdisciplinary research on Human-Machine-Interface (HMI)

It is important to pass on information from a driver assistant system to the driver in an optimal way, thereby considering his psychological and physiological characteristics. In order to meet this challenge, research on driver assistant systems is done interdisciplinary within the I-FAS.

Therefore, the driving behavior is observed within the Carai, for example by means of various eye trackers. Practical application can basically be found in every Human Machine Interface (HMI) within a vehicle.