Jump label

Service navigation

Main navigation

You are here:

Main content

MoFFa

MoFFa: Holistic model to describe the allocation and the transfer of tasks between the human driver and Advanced Driver Assistance Systems during automated and interconnected driving 

Funded by the federal ministry of TRANSPort and DIGITAl INFRASTRUCTURE

Prof. Dr. Johannes Weyer, Marco Hellmann, TU Dortmund
Duration: 01.07.2017 - 30.06.2020
In Cooperation with the Institute of Control Theory and Systems Engineering (RST) at TU Dortmund and the Research and Technology Centre “Ladungssicherung Selm gGmbH” (LaSiSe)

 

Research context 

Cars become increasingly digital and automated by Advanced Driver Assistance Systems. While humans were responsible for nearly all driving tasks in the past, today more and more work is redistributed to technology. Therefore, new mechanisms are needed to constitute a clear allocation of responsibilities and coordination of interactions between the human driver and Assistance Systems within the vehicle:

  • When are humans able or allowed to take over certain driving tasks or should the tasks rather be handed over to the car’s assistance system?
  • How do humans get informed when there is a need to take over driving tasks?
  • How does the handover have to take place so that the human accepts the assistance technology and willingly uses it without the formation of stress?
  • To answer these questions, a holistic model describing mechanisms of task handover and takeover mathematically is developed. 

The model is based on the “levels of automation” proposed by the Society of Automotive Engineers (so-called SAE levels) and is adapted for a systematic analysis of functions for automated and interconnected driving. Moreover the MoFFa project is supposed to

  • observe and coordinate the status of the human driver and the Assistance System
  • recognize and analyze problematic constellations of the interaction between the human driver and the Assistance System in different SAE levels
  • describe automated driving functions exactly with regards to transparency towards users and conformance to user expectations
  • provide a gain of knowledge concerning safety, social acceptance as well as user-focused acceptance for automated driving

Research methods

The holistic model is developed in a three-step process using different research methods (quantitative online-questionnaire, experiments and interviews).

  1. Firstly, an online-questionnaire about driver-vehicle-interactions is used to classify different types of drivers and to identify predictor variables that have an impact on the state of the driver during vehicle use (e.g. stress, experience, attention). The questionnaire helps to construct an instrument that is able to match test subjects with specific types of drivers at a later stage of the project. As a result, different reactions of different driver types when handing or taking over driving tasks can be observed.

  2. Secondly, a static driving simulator is used to optimize the description of interactions and interdependences as part of the holistic model. Therefore test subjects that use the driving simulator are exposed to "normal" and "critical" driving situations in the simulator. Additionally the test subjects participate in a pre-post-survey which examines, for example, user acceptance of specific task handovers or takeovers.

  3. Thirdly, the holistic model based on the driving experiments is evaluated by exposing test subjects to drive in real situations. Afterwards, results with the ones under laboratory conditions (driving simulator) are compared and interpreted. The experiment takes place on the test route of the “Forschungs- und Technologizentrum Selm gGmbH”. As a last step, user acceptance of task transfer is again analyzed by doing a pre-post-survey with different driver types.