![](bg6.png)
application. They showed that their model is dead-
lock-free and conflict-free. In our case, Petri Nets
are used instead as a high level formalism for the
MDP, which derives the strategy of the co-pilot
itself and the intelligence of the automated system.
To sum up, in this position paper we have at-
tempted a new technological system for co-pilot
implementations, using MDP for the computational
implementation of the cognitive system. Since
HOLIDES project have just started October 2013,
the development of the proposed framework is in its
initial phase. The next steps consist now in the prep-
aration and execution of the experimental phase on
the field with the demonstrator vehicle to collect
real-time and on-line data for the tuning and the
evaluation of the MDP co-pilot. This phase, together
with the prototype set-up, is foreseen in this year and
at the beginning of 2015; while the final implemen-
tation and assessment of the co-pilot will be the
activity to carry out within the end of the project
(August 2016).
One important future work will be to investigate
the possibility of extending this approach using
Partially Observable MDP (POMDP) (Leslie,
1995)). Indeed, a POMDP is a generalization of an
MDP, in which an agent must base its decisions on
incomplete information about the state of the envi-
ronment. Hence, POMDP can be used more effi-
ciently to model systems where the agent cannot
directly observe the complete underlying state.
However, POMDPs are often computationally in-
tractable to solve a real system and its approximate
solution techniques for POMDPs could not provide a
sub-optimal solution that satisfies the time con-
straints imposed by our application.
In addition, another important achievement is
represented by the full exploitation of the CASCaS
framework, in particular for the cognitive behaviour.
In this context, the integration of driver’s state clas-
sifier inside the co-pilot (driver state becomes an
input in this case) is a crucial point for deciding the
level of automation.
ACKNOWLEDGEMENTS
This research has been performed with support from
the EU ARTEMIS JU project HoliDes
(http://www.holides.eu/) SP-8, GA No.: 332933.
Any contents herein reflect only the authors' views.
The ARTEMIS JU is not liable for any use that may
be made of the information contained herein.
REFERENCES
Tango, F., and Botta, M. (2013). Real-Time Detection
System of Driver Distraction Using Machine Learn-
ing. IEEE Transactions on Intelligent Transportation
Systems Journal, vol. 14, no. 2, pp. 894-905, DOI
10.1109/TITS.013.2247760
Inagaki, T. (2008). Smart collaboration between humans
and machines based on mutual understanding. Annu.
Rev. Control, vol. 32, no. 2, pp. 253–261.
Flemisch, F. O., Adams, C. A., Conway, S. R., Goodrich,
K. H., Palmer, M. T., and Schutte, P. C. (2003). The
H-Metaphor as a Guideline for Vehicle Automation
and Interaction. Control, no. December, pp. 1 – 30.
Heide, A. and Henning, K. (2006). The ‘cognitive car’: A
roadmap for research issues in the automotive sector.
Annual Reviews in Control, vol. 30, no. 2, pp. 197–
203.
Leslie Pack Kaelbling, Michael L. Littman, and Anthony
R. Cassandra. Planning and acting in partially observ-
able stochastic domains. Artificial Intelligence,
101:99-134, 1995.
Li, L., Wen, D., Zheng, N.-N., and Shen, L.-C. (2012).
Cognitive Cars: A New Frontier for ADAS Research.
IEEE Transactions on Intelligent Transportation Sys-
tems Journal, vol. 13, no. 1, pp. 395–407.
Salvucci, D. D. (2006). Modeling driver behavior in a
cognitive architecture. Human Factors, 48, pp. 362-
380.
Lüdtke, A. Weber, L., Osterloh, J.-P., Wortelen, B.
(2009). Modeling Pilot and Driver Behaviour for Hu-
man Error Simulation. Conference Proceedings, HCI
2009, Digital Human Modeling, San Diego, Springer:
LNAI 2009.
Moore, R., and Lopes, J. (1999). Paper templates. In
TEMPLATE’06, 1st International Conference on
Template Production. SCITEPRESS.
Smith, J. (1998). The book, The publishing company.
London, 2
nd
edition.
NaiQi Wu, Feng Chu, Mammar, S., MengChu Zhou
(2011). Petri Net Modeling of the Cooperation Behav-
ior of a Driver and a Copilot in an Advanced Driving
Assistance System. In Intelligent Transportation Sys-
tems, IEEE Transactions on (Volume:12 , Issue: 4 ).
Date of Publication: Dec. 2011.
M. Beccuti, G. Franceschinis, and S. Haddad. Markov
Decision Petri Net and Markov Decision Well-Formed
Net Formalisms. In 28th International Conference on
application and theory of Petri nets and other models
of concurrency (ICATPN'07), volume 4546 of Spring-
er LNCS, pages 43-62. June 25-29, 2007, Siedlce, Po-
land.
ICINCO2014-11thInternationalConferenceonInformaticsinControl,AutomationandRobotics
822