Human-Machine Interaction for Urban Environments

Design of the human-machine interface (HMI) has been investigated in several previous projects, but with a focus on a specific solution for each particular function. This strategy is no longer adequate for urban driving: With an increasing number of functions competing for the attention of the driver, lack of coordination can result in overburdening. Moreover, in urban areas, the probability of competing alerts, warnings, or interventions is higher than on freeways or country roads. This problem motivates the following three aims of the sub-project:

Aim 1: Conception and development of a generic HMI toolkit

The multi-functional, integrated HMI toolkit for urban areas will be created by structuring, standardizing and adapting HMI concepts for assistance and safety functions that already exist or are currently being developed. The toolkit includes a strategy for systematic derivation of action-oriented HMI concepts for applications ensuring safe, comfortable and efficient driving. The main idea is that, even if a particular required action (such as “braking”) might be triggered by various events, the system must behave consistently and predictably regarding information, warnings and interventions leading to the action. To this end, the HMI toolkit provides an interface for controlling and prioritizing HMI output media. Strategies for timing combinatorics, as well as for phased escalation of sensory input modes to the driver (i.e., visual, tactile, and acoustic) are also included in the toolkit.

The strategic approach of the HMI toolkit is oriented toward requirements of urban areas; user-specific characteristics (such as performance of older drivers) and future requirements for individual mobility are important elements of this approach. Integration of navigation information and environmental knowledge will allow optimization of warning modes and timing and support anticipatory driving. For example, hazards not visible to the driver can be indicated by the use of tactile feedback on the accelerator pedal and the steering wheel. This feedback mode not only reduces risks, but also leads to an anticipatory driving style that reduces stress and contributes to significant decreases in fuel consumption and emissions, particularly for trucks.

A key challenge is to convey the vehicle’s “information advantage” to the driver in a safe and useful manner: The information must be reliable, and the presentation must be comprehensible.

Aim 2: Integration of UR:BAN Applications

Testing of the HMI toolkit logic and the integrative HMI concepts derived from this logic will be based on functional specifications from the projects “Cognitive Assistance” and “Networked Traffic System”. The focus is on the sub-project applications "Collision Avoidance by Swerving and Braking", "Protection of Vulnerable Road Users" and "Safe Transverse and longitudinal Control in Cities" and the sub-projects “Smart Intersection”, “Urban Roads” and “Regional Network”. Information and warnings from the sub-project “Behavior Prediction and Intention Detection”, which do not involve direct intervention in driving behavior, will be tested as well.

Aim 3: Test and Demonstration of the HMI Solutions

Throughout the entire development process, the resulting HMI concepts will be evaluated from a user perspective using appropriate methods, including model-based simulation, expert testing, and direct experimentation. Beginning with preliminary studies, the sub-project will first obtain refined requirements; it will then carry out an iterative development and evaluation process and continue up to evaluation of the overall approach. The focus is on optimized design and investigation of classical usability issues.