Vehicles with highly automated driving (dt. „Hochautomatisiertes Fahren“; SAE Level 3) functionalities take over longitudinal and lateral control for a certain period of time in a specific situation without the need for monitoring by the driver (Gasser et al., 2012). Highly automated driving is no longer a distant future scenario. The first highly automated systems were already approved for public roads in the United States in 2015 (Daimler AG, 2018). There should be no technical obstacles to introducing Highly automated driving on German roads by 2020 (Bengler et al., 2014). In addition to increased safety, efficiency, and comfort (Bengler et al., 2014), the possibility of carrying out NDRTs is expected to increase productivity (Kaur and Rampersad, 2018). Professional drivers can recover during automated driving or carry out planning activities such as route planning (Flämig, 2015).
When system limits are reached, or errors occur, the human driver must regain vehicle control with sufficient time reserves (Gasser et al., 2012, p. 31). This change of control between humans and vehicles is a critical factor in highly automated driving. Problems and dangers are caused by inadequate trust, insufficient situational awareness, and loss of skill (Manzey, 2012).
Human-Machine Interface Concept for Highly Automated Truck Driving
To meet the resulting challenges, a suitable human-machine interface (HMI) for highly automated truck driving is required. The majority of research work on highly automated driving is currently focused on passenger cars. The knowledge gained from this provides a sound basis but cannot be transferred directly to trucks (Richardson et al., 2018). As part of this master’s thesis, a developed HMI prototype (based on: Czaplarski, 2018; Lehmer, 2018; Richardson et al., 2018) for highly automated driving in trucks is evaluated and optimized human-centered.
Evaluation
Simulator Study
In a driving simulator study with 30 test subjects, the truck drivers experience the highly automated system in a dynamic truck driving simulator. Six truck-specific highway scenarios (Richardson et al., 2017) occur in the simulation. In four of the six scenarios, the driver must take over, whereby in two situations each, a takeover request is made early (60 s beforehand) and immediately (10 s beforehand). The drivers are busy reading a magazine article during the highly automated driving, so they are entirely distracted when the system prompts them to take over. The participants make three trips with varying information on the remaining highly automated driving availability. Depending on the variant, the HMI provides information in the form of time or distance details or no information. During and after the journeys, the users evaluate the variants regarding controllability, workload, acceptance, user experience, information content, and well-being. All variants receive good ratings.
The two variants with information on the remaining highly automated driving availability achieve significantly better results than the one without information. This indicates a positive effect of the information provided on the interaction between the driver and the automation. These results are reflected in the qualitative final survey, in which the understanding of the system and the evaluation of individual components and information units, as well as aspects of the overall concept, are recorded. The final survey is rounded off by a „mix and match“ design workshop in which the participants design their own display concepts and express ideas and suggestions.
Assessment of Conformity with ISO 15005
In addition to the driving simulator study, the HMI is evaluated based on inspections concerning conformity with ISO 15005. Most of the requirements and recommendations of the nine dialog principles have already been fulfilled by the HMI. However, there are significant deficits with regard to the principles of compatibility with vehicle guidance, timing/priorities, consistency, and conformity with driver expectations.
Optimization Potential and Further Steps
The findings from the driving simulator study and the conformity assessment are bundled in the form of optimization measures (see Thesis Chapter 6.2), on the basis of which specific optimization proposals are developed (see Thesis Appendix D). The main adjustments concern the consistent design of the displays across all modalities and components, the increase in the level of detail of information units such as the map display, the expansion of the warning cascade to include a haptic modality, and the display of post-hoc information after a takeover.
The work lays the foundation for the iterative further development of the HMI. In line with the human-centered development process, the revised interface should be re-evaluated with users. In particular, critical scenarios with required driver reactions and the long-term use of the system should be investigated.
References
K. Bengler, K. Dietmayer, B. Färber, M. Maurer, und H. Winner, „Three Decades of Driver Assistance Systems – Review and Future Perspectives“, IEEE Intell. Transp. Syst. Mag., Bd. 6, Nr. 4, S. 6–22, 2014.
M. Czaplarski, „Umsetzung und Evaluation einer Mensch-Maschine-Schnittstelle für das hochautomatisierte Fahren im Lkw“, Masterarbeit, Technische Unviersität München, 2018.
Daimler AG, „Autonom durch Nevada. Freightliner Inspiration Truck“, 2018. [Online]. Verfügbar unter: https://www.daimler.com/innovation/autonomes-fahren/freightlinerinspiration-truck.html. [Zugegriffen: 11-Sep-2018].
H. Flämig, „Autonome Fahrzeuge und autonomes Fahren im Bereich des Gütertransportes“, in Autonomes Fahren, M. Mauerer, J. C. Gerdes, B. Lenz, und H. Winner, Hrsg. Wiesbaden: Springer Vieweg, 2015, S. 377 – 398.
Fraunhofer IAO, „Hochautomatisiertes Fahren auf Autobahnen – Industriepolitische Schlussfolgerungen“, 2015.
T. M. Gasser, C. Arzt, M. Ayoubi, A. Bartels, L. Bürkle, J. Eier, F. Flemisch, D. Häcker, T. Hesse, W. Huber, C. Lotz, M. Maurer, S. Ruth-Schumacher, J. Schwarz, und W. Vogt, „Rechtsfolgen zunehmender Fahrzeugautomatisierung“, Bergisch Gladbach, 2012.
K. Kaur und G. Rampersad, „Trust in driverless cars: Investigating key factors influencing the adoption of driverless cars“, J. Eng. Technol. Manag., Bd. 48, S. 87–96, 2018.
C. Lehmer, „Konzeption und Evaluation einer Mensch-Maschine-Schnittstelle für das hochautomatisierte Fahren im Lkw“, Masterarbeit, Technischen Universität München, 2018.
D. Manzey, „Systemgestaltung und Automatisierung“, in Human Factors, 2. Aufl., P. Badke-Schaub, G. Hofinger, und K. Lauche, Hrsg. Berlin Heidelberg: Springer, 2012, S. 333–352.
N. Richardson, B. Michel, A. Zimmermann, und F. Diermeyer, „Erfassung und Bewertung des Informationsbedarfs von Lkw-Fahrern während hochautomatisierter Fahrt“, in 9. VDI-Tagung – Der Fahrer im 21. Jahrhundert, 2017, S. 1–13.
N. T. Richardson, C. Lehmer, B. Michel, und M. Lienkamp, „Conceptual design and evaluation of a human machine interface for highly automated truck driving“, Intell. Veh. Symp., S. 2072–2077, 2018.
Metadata
Master’s Thesis
Advisors & Reviewers
Institutions
Related Publications
Natalie T. Richardson, Lukas Flohr, and Britta Michel. 2018. Takeover Requests in Highly Automated Truck Driving: How Do the Amount and Type of Additional Information Influence the Driver–Automation Interaction? Multimodal Technologies and Interaction 2, 4 (2018), 68. https://doi.org/10.3390/mti2040068
Natalie Tara Richardson. 2020. Konzeption und Langzeittest der Mensch-Maschine-Schnittstelle für das hochautomatisierte Fahren im Lkw. Dissertation. Technical University of Munich, Munich, Germany. https://mediatum.ub.tum.de /1520460