- Traffic psychology
- Human factors
- Geography and traffic planning
- Multi-sensor systems and distributed computing
The aim of a user-centered research approach is to focus on the user during all phases of the development and evaluation. Therefore we follow an iterative process from the first ideas to the evaluation of the prototype.
Older pedestrians are often involved in traffic incidents. Due to age-related declines they are not able to pay sufficient attention to the ongoing road traffic. Besides age-related deficits in perception and motion execution, the reason for this can particularly be found in deficits in pre-attentive and attentive processes. These deficits can hinder the phase of approaching roads and with that the decision of when, where and how to cross a street. Therefore, one aim of the FANS research group is to analyse road crossing behaviour of older pedestrians. On one hand, we want to find out whether older pedestrians are aware of their deficits and on the other hand whether they develop appropriate strategies to compensate the deficits. A combination of different methods such as observations, interviews, and questionnaires will be used to address these questions. In addition, the thus obtained results will be compared with those of younger pedestrians to draw conclusions about the specific situation of the older pedestrians.
Efficient and safe movement executions often claim high concentration and cognitive control of older people. For targeted and successful movements, age-related deficits of motor functioning and sensory processing must be compensated for, often requiring the usage of aids. Thus, former automated motion sequences become challenges needing extra focus of attention and widened control mechanisms. In addition, planning and executing movements can interfere with simultaneously executed secondary tasks such as making telephone calls or observing traffic. The project aims to systematically investigate age-related changes in cognitive perception and processing as well as loss and slowdown of motor responses. Records of superficial brainwaves (Electroencephalography, EEG) will be analysed to gain more knowledge of information processing of different stimuli modalities (visual and tactile). The obtained data will be completed by an analysis of walk and posture parameters. Data will be collected under controlled laboratory conditions and validated in realistic movement situations in the Berlin Mobile Brain/Body Imaging Laboratory (BeMoBIL) of the Department of Biological Psychology and Neuroergonomics at TU Berlin. The aim of the project is to get a better understanding of perception and processing of visual and tactile stimuli and of possible interdependencies with simultaneously executed motion sequences.
The research group FANS aims to support older people crossing roads safely compensating for their attentional deficits. A technical system will be used to draw their attention to potential dangerous situations. This part of the project is about analysing and comparing feedback modalities being used to direct attention. In a simulated street environment, different feedback modalities such as vibration, thermic or auditive signals will be compared regarding their effectiveness in directing gaze. Eye-tracking analysis will help to show how reliable and under what circumstances these feedback modalities can direct gaze to locations/situations where target stimuli may appear.
The DFG-project „Investigating the impact of alarm systems with adaptable thresholds in younger and older adults on the example of a pedestrian assistance system“, which is affiliated to the FANS research group, focusses on more fundamental research questions. By conducting several experiments, using the same video based simulation environment, the advantages of alarm systems with adaptable thresholds for younger and especially older people will be investigated.
Geography and traffic planning
For older people the preservation of everyday mobility is an important aspect of living an independent and self-determined life in order to accomplish everyday tasks, such as shopping or using medical services. However, everyday mobility is also a basic social need for perceiving oneself as part of the society and interacting with other people. Not being able to engage in social exchanges, because of mobility barriers, can lead to isolation. Due to the fact that older people in urban agglomerations cover most of their daily ways by foot, the focus of the research group FANS lays on mobility barriers in terms of older pedestrians. In this context, the physical and cognitive losses of older people as well as structural factors of the environment will be taken into account; their relations to each other will be examined, and compared to the current legal and planning guidelines. The findings obtained will be used to propose a theoretically sound, integrated and actor-centered traffic development concept that increases the everyday mobility of older people incorporating a pedestrian assistance system. The traffic concept will be based on the results of several types of interviews and observations. In addition, secondary statistics will be processed and analyzed to ensure a holistic view on the problem.
Multi-sensor systems and distributed computing
The aim of this part of the project is the environment recognition and orientation from the perspective of older pedestrians. Therefore, a multi-sensor approach is used. The idea is to equip a walking frame with several sensors (e.g. camera, distance sensors, GPS, inertial measurement unit, etc.) providing a reliable representation of the actual situation. As hardware for the control and data processing, a single board computer with a microcontroller is used. In a first step, it will be tested whether different types of sensors already successfully used in Advanced Driver Assistance Systems and robotics are suitable for our purpose. As a result, we will determine the parameters of the sensors and derive possible use cases. Afterwards we will develop a knowledge-based system as a framework that will manage the sensors modularly, flexibly, and robustly as agents. For data fusion, we will implement machine learning and data mining methods.