Information Technology Reference
In-Depth Information
- The V3 cycle, which is longer still as it requires the use of the simulator. It
enables the integration of the assistance to be validated, in particular for
communication with the driving simulator, thanks to the integration platform which
functions in “online” mode (that is to say on the data transmitted in real time) and
intermediary evaluations of different versions of the assistance in critical scenarios
to be carried out.
Development of the
assistance
Implementation
Integration of the assistance
Continuous Integration
V2 : Integration platform (replay
offline)
V3 : Integration platform (online)
Definitive
assistance
systems and HMI
Specifications of
the assistance
Driver model
HMI environment mo0del
Specification
of the
D
assistance
Final
ergonomic
evaluation
Integrated
assistance
system
Assistance/
Environment
communication
V1
Feedback
expert
Technical
constraints
V2
V3
Tests of the assistance
Functional validation
Legend
Activity
Objectives
Tools/Methods
V1 : Scenario simulation
V2 : Integration platform (replay
offline)
Intermediary evaluation
of the assistance
Production
Figure 7.9.
Development process of the VIVRE2 system according to a UCD approach
7.5.2.
Description of an adaptive solution
The architecture of the system was developed by using an “adaptive
technological” type approach, i.e. that adapts the functioning of the system to take
into account the behavior of the driver in a given situation [BEL 05]. For this, a
model of the intention of the driver was implemented. This model is used by the
assistance like any input on a par with the data of the perception of the external
environment.
7.5.2.1.
Vulnerable user detection zones
The system selects the type of assistance with the help of two available
information sources: information on the possible position of vulnerable users (in
front, behind, or on the right-hand side of the vehicle) and information regarding the
actions of the driver (Figure 7.10). This information is then compared to the
intentions of the driver, which are inferred from a “driver intention model”, in order
to determine the adapted assistance strategy, and then to run the HMI.
