Autonomous Vehicles - Relations to Human Intelligence Lehrstuhl fr - PowerPoint PPT Presentation

Lehrstuhl für Ergonomie Technische Universität München Autonomous Vehicles - Relations to Human Intelligence Lehrstuhl für Ergonomie Technische Universität München Prof. Klaus Bengler 13.11.2015

Lehrstuhl für Ergonomie Technische Universität München …. We need data for the HMI design for The support of mode awareness The anticipation To increase trust To prepare transitions and arbitration This is Prospective level of service Identification of objects Reproducable behavior 13.11.2015 2

Lehrstuhl für Ergonomie Technische Universität München

Lehrstuhl für Ergonomie Technische Universität München Human Intelligence .. increases and decreases traffic safety (getting ccoperative vs. bending the rules) For automation human intelligence is something • to be replaced („Take the human out of the loop !“) • to be supported („ We still have to work on this “) • to learn from („ How to understand a scene ?“)

Lehrstuhl für Ergonomie Technische Universität München Automation in a Human-Machine-System • Should not loose performance in summary • There will still be humans all over the place • These humans will change their behavior

Lehrstuhl für Ergonomie Technische Universität München Models of Human Behavior and HMI from: Winner, H., Hakuli, S. & Wolf, G. (2012). Handbuch Fahrerassistenzsysteme

Lehrstuhl für Ergonomie Technische Universität München Take-Over in Automated Vehicles Take-Over (here): Machine initiated transition from automated to manual driving. Take-Over Request (TOR) System Limit Time Budget Manual Driving Highly Automated Driving Take-Over / Transition Non-Driving Reaction Related Task Time

Lehrstuhl für Ergonomie Technische Universität München Information Processing Wickens, C. & Carswell, C. (2006). Information Processing. In G. Salvendy (Hrsg.), Handbook of Human Factors and Ergonomics (S. 111-149). Hoboken: John Wiley.

Lehrstuhl für Ergonomie Technische Universität München To Replace Human Intelligence Benefits in the area of • very short reaction times, situative automation • permanently loading and monotonous tasks • reduced human skills and info-processing abilities

Lehrstuhl für Ergonomie Technische Universität München Lack of Arousal or to Much Arousal Yerkes-Dodson-Law (1908): performance and earousal Scenario: Scenario: „Support in urban traffic “ „Relief on Motorways “ „ Compensating driver „Automation in congestion distraction “

Freedom – Drivers will „ Learn “ the End of Distraction will they „ Learn “ the levels of automation or develop an own model? UR:BAN • UR:BAN Plenum 2015 Impulsvortrag • Endlich Freiheit für die MMI! • 03.03.2015 • Dresden

Lehrstuhl für Ergonomie Technische Universität München Controllability / Visual Information Preview of system behavior leads to Quicker responses and better learnability by visualisation But still to total misses of traffic signs while automated driving (Weißgerber 2012)

Lehrstuhl für Ergonomie Technische Universität München Visual Behavior During Tertiary Task (Damböck 2012) Mean Single Glance Time ± 95%-CI [s] Mean Glance Frequency ± 95%-CI [1/s] 2,5 0,7 0,6 2,0 0,5 1,5 0,4 0,3 1,0 0,2 0,5 0,1 0,0 0,0 manuell assistiert teilautom_mH teilautom_oH manuell assistiert teilautom_mH teilautom_oH Szene Tacho CID Szene Tacho CID N=24 N=24 • Higher LOA lead to more visual divertion • Haptic information is used to reducte glances to speedometer • Visual behavior can be used as an availability metric for cooperation [H-M-S] (Damböck 2012)

Lehrstuhl für Ergonomie Technische Universität München To Replace Human Intelligence High/Full automation systems • Have to be highly reliable and trustable • Should know about user state and intention • Have to be learnable to enable intelligent usage patterns and cooperative behavior • The systems adapt to surrounding traffic or serve as good example?

Lehrstuhl für Ergonomie Technische Universität München Automation to Support Human Intelligence In areas of • Weaknesses on both sides • Interesting, experiencing activities

Lehrstuhl für Ergonomie Technische Universität München Automation to Complete and Support Human Intelligence Shared control systems • Have to be cooperative • Have to be transparent for the user • Could be user adaptive • Will be an interesting mode between automation and manual driving

Lehrstuhl für Ergonomie Technische Universität München Automation to Learn From Human Intelligence Automation could benefit from knowledge about • Situation understanding of drivers • Generation and application of driving strategies • Cooperative problem solving • Learning of complex sceneries

Lehrstuhl für Ergonomie Technische Universität München To Drive Safely is a Learning Process Accicents and Age, 2008 80 Proportion of Causation in % 70 60 50 40 30 20 10 0 Quelle: Statistisches Bundesamt, all gesamt männlich weiblich Verkehrsunfälle 2008

Lehrstuhl für Ergonomie Technische Universität München Humans Situation Interpretation ExampleTailgate Inattentive driver, braking late (Video: Inattentive driver, braking late)

Lehrstuhl für Ergonomie Technische Universität München SEEV-Modell as an Example The probability of a glance is a function of Salience, Effort, Expectancy and Value

Lehrstuhl für Ergonomie Technische Universität München Environment and Perception The traffic environment is and will be optimized for human perception This trained our perception: • Signs, lines • Indicators • Wheels • Trajectories

Lehrstuhl für Ergonomie Technische Universität München Automation Means • Users will change their behavior due to positive or negative experience (i.e. learning) • Surrounding traffic will change its behavior

Lehrstuhl für Ergonomie Technische Universität München Research on • Longterm usage in „normal“ situations • Behavorial styles in mixed traffic • Mental models of users about system modes • Relevance of driver status and intention • Cooperation (Car<-> user, Car2Car) • Evaluation of adaptive/learning automation

Lehrstuhl für Ergonomie Technische Universität München References • 2008. FISITA World automotive Congress. • Abbink, D. A., Mulder, M., & Boer, E. R. (2012). Haptic shared control: smoothly shifting control authority? Cognition, Technology & Work, 14 (1), 19 – 28. doi:10.1007/s10111-011-0192-5 • Ardelt, M., Coester, C., & Kaempchen, N. (2012). Highly Automated Driving on Freeways in Real Traffic Using a Probabilistic Framework. IEEE Transactions on Intelligent Transportation Systems, 13 (4), 1576 – 1585. doi:10.1109/TITS.2012.2196273 • Badke-Schaub, P., Hofinger, G., & Lauche, K. (2012). Human Factors: Psychologie sicheren Handelns in Risikobranchen (2., überarbeitete Auflage). Heidelberg: Springer. Retrieved from http://dx.doi.org/10.1007/978-3-642-19886-1 • Banks, V. A., Stanton, N. A., & Harvey, C. (2014). What the drivers do and do not tell you: using verbal protocol analysis to investigate driver behaviour in emergency situations. Ergonomics, 57 (3), 332 – 342. doi:10.1080/00140139.2014.884245 • Beggiato, M., & Krems, J. F. (2013). The evolution of mental model, trust and acceptance of adaptive cruise control in relation to initial information. Transportation Research Part F: Traffic Psychology and Behaviour, 18, 47 – 57. doi:10.1016/j.trf.2012.12.006 • Beller, J., Heesen, M., & Vollrath, M. (2013). Improving the Driver-Automation Interaction: An Approach Using Automation Uncertainty. Human Factors: The Journal of the Human Factors and Ergonomics Society, 55 (6), 1130 – 1141. doi:10.1177/0018720813482327 • Carsten, O., Lai, F. C. H., Barnard, Y., Jamson, A. H., & Merat, N. (2012). Control Task Substitution in Semiautomated Driving: Does It Matter What Aspects Are Automated? Human Factors: The Journal of the Human Factors and Ergonomics Society, 54 (5), 747 – 761. doi:10.1177/0018720812460246 • Damböck, D., Farid, M., Tönert, l., & Bengler, K. (Eds.) 2012. Untersuchung von Einflüssen automatischer Bremsmanöver und Verkehrssituationen auf die Übernahmezeit und -qualität in hochautomatisierten Fahrzeugen. • Damböck, D., Weissgerber, T., Kienle, M., & Bengler, K. (2013). Requirements for cooperative vehicle guidance. Proceedings of the 16th International IEEE Annual Conference on Intelligent Transportation Systems (ITSC 2013), 1656 – 1661. doi:10.1109/ITSC.2013.6728467 • Damböck, D. (2013). Automationseffekte im Fahrzeug - von der Reaktion zur Übernahme (Dissertation). Technische Universität München, München. • Damböck, D., Farid, M., Tönert, L., & Bengler, K. (2012). Übernahmezeiten beim hochautomatisierten Fahren. 5. Tagung Fahrerassistenz . Retrieved from http://www.ftm.mw.tum.de/uploads/media/24\_Damboeck.pdf