111 / 308
111
Ergonomics
In addition, new concepts of dividing the driving task
between the machine and the human become the
center of attention for the ongoing research. Finally, the
human-centered approach of designing automated driving
functions raises issues of how to take into account every
driver’s individual preferences on the road.
@City – Automated Cars and Intelligent Traffic in the City
will be conceived to support the driver in complex urban
settings, which are characterized by a high information
density and short response time. The demonstration
of these automated driving functions will be realized in
exemplary pilot applications.
PAKoS – Personalized, Adaptive, Cooperative Systems
for Automated Vehicles
The project @City funded by the Federal Ministry for
Economic Affairs and Energy started in September 2017
and will run for four years.
Due to the complexity of urban traffic areas cities set
some special challenges for use cases in the automated
driving context. Various roundabouts and crossings
complicate the situation understanding compared with
a well-structured environment like motorways. Reliable
situation understanding of the setting lays the foun-
dation for every automated driving function. The goal
of the project @CITY is to enable automated driving in
urban areas. Automated vehicles and intelligent traffic
are the key elements for safe, efficient and stress-free
future traffic. They should not only provide the highest
possible assistance, but also enhance the interaction
between vehicle and driver or rather between vehicle
and vulnerable road users like pedestrians and cyclists.
The project aims at efficient and robust algorithms for
environment detection and the driver’s situation under-
standing. Moreover, its goals are the achievement of a
precise digital map, well-designed automation as well
as the ideal integration of the human driver. Based on
these goals automated driving functions for urban areas
(Reference: Consortium @City, 2017)
In three years’ time, the project PAKoS aims at introducing
an automation manager capable of displaying the full use
case of an automated vehicle.
This process starts with the possibility for the driver to
prepare the automated vehicle remotely via a smartphone
app in a way that it already suits the driver’s personal
preferences upon arrival. During the drive the aim is to
monitor and identify the user state and then to use this
information to adapt the automated driving functions
appropriately. The third big part of PAKoS focuses on a
new way of handling the take-over process during which
the driver has to regain full control of the vehicle. Unlike
the conventional binary switch from automation to driver, a
cooperative approach is investigated, in which automation
and driver simultaneously control the vehicle during the
take-over process.
To tackle these interdisciplinary challenges, a consortium
of nine German partners was formed: Karlsruher Institute
of Technology, Technical University of Munich, Fraunhofer
Institute of Optronics, System Technologies and Image
Exploitation, Robert Bosch GmbH, BMW AG, Spiegel
Institute Mannheim GmbH & Co. KG, Videmo Intelligent
Video Analysis GmbH & Co. KG, mVise AG, b.i.g. security.
(Reference: Consortium PAKoS, 2016)
3. Interaction & Cooperation
The research group investigates interaction processes
between a human agent and one or multiple cooperative
partners. These partners can represent other humans as
well as technological devices such as robots, vehicles
or aircraft. In the context of cooperation, it is important
to balance partially competing individual and shared
goals of the parties involved, which can be achieved
by diverse mechanisms of communication. Metrics are
to be developed to allow for the performance-oriented
assessment of cooperation across multiple domains. With
regard to robots, effective strategies are required for both
locomotive and evasive actions by designing legible and
predictable movements. A robot’s motion can implicitly
convey intentions to the human agent, who, visually per-
ceiving and interpreting the respective information, reacts
by performing certain movements themselves within