Transparency in highly automated shuttles – How human-machine interfaces can strengthen understanding and trust
- Highly automated vehicles are revolutionising the world of mobility, but without people at the wheel they present passengers with new challenges.
- Safety manoeuvres require clear communication between the vehicle and passengers.
- Transparency about the perception and decisions of AI should strengthen trust and acceptance.
- Findings from studies: Targeted information increases understanding, improves the experience for users and makes highly automated mobility solutions more accessible and trustworthy.
- Focal points: Automation, artificial intelligence, human-machine communication
Highly automated vehicles (HAVs) will revolutionise the mobility landscape in the future. However, the introduction of HAVs presents passengers with new situations. Without people at the wheel, they will have to rely on artificial intelligence (AI) and remote operators. Uncertainty is particularly high in the case of minimal risk manoeuvres (MRMs) - safety manoeuvres that the vehicle performs in the event of incidents that cannot be resolved directly by the system. Such situations often include everyday events such as leaves on the road or more complex scenarios such as road accidents. The legal requirements, such as the AFGBV, stipulate that the vehicle stops driving in such cases and waits for support from remote operators.
These situations require clear communication between the automated system and the passengers, which should be achieved by means of modern human-machine interfaces (iHMI). The iHMI should create transparency with regard to the decisions made by the AI and thus strengthen the trust and acceptance of the users. The aim of a study by the German Aerospace Center (DLR) was therefore to increase understanding of vehicle behaviour and thus strengthen trust in the technology.
The challenge: uncertainty and complexity in HAVs
But how do HAVs make their decisions? They use AI to analyse their environment and calculate the best alternative course of action. The vehicles' decision-making processes are often difficult to understand if the system is not transparent and does not provide any information about the decision-making process and remote operators have to be involved. These complex processes are often incomprehensible to passengers, which can lead to a feeling of insecurity. And this is where the idea of transparency comes in: By providing targeted information about how the vehicle perceives and makes decisions, the aim is to facilitate interaction with the technology and thus increase trust in it.
Approaches to improving transparency
The study investigated how different iHMI designs can improve the passengers' understanding of the automation processes. The study focussed on two central factors:
- Observability: Information about what the vehicle "sees" (see Fig. 1).
- Reasoning: Explanations of why the vehicle makes certain decisions together with the remote operator actively involved in the solution and the logic behind them.
The study showed that passengers develop a better understanding of the system through these two types of information. At the same time, the predictability of the vehicle's behaviour was significantly improved. This led to a more positive experience for users and increased confidence in the previously unknown technology. Another important aspect is the adaptability of the information provided. Different situations and user groups require different levels of transparency. The conclusion from this realisation is that an iHMI should not only be transparent, but also flexible.
Results: More transparency, better user experience
The results of the study provide valuable insights for the design of iHMIs in HAVs:
- Improved understanding and trust: Greater transparency leads to a better understanding of vehicle processes and strengthens the trust of passengers.
- Stable usability: Greater transparency does not impair the usability of the system. However, there are indications that information overload could be possible if transparency is too high.
- Demand-orientated design: Passengers benefit from the provision of information tailored to their individual needs and situation.
It is particularly important to emphasise that transparency can have a positive effect on the trust and acceptance of HAVs without compromising user-friendliness. This makes transparency a key design element for future automated mobility solutions.
Future prospects: transparency as the key to acceptance
These findings lay the foundation for further research. Future studies should determine the ideal level of transparency in order to achieve a balance between a wealth of information and comprehensibility. It is also necessary to validate the results in real environments or driving simulators in order to test the applicability of the findings. Another aspect of research is the user-friendly presentation of remote operation processes in order to promote trust and understanding in critical scenarios. Furthermore, it remains to be clarified how long-term experiences with highly automated vehicles can be positively influenced by transparency.
The introduction of highly automated shuttles is not only changing mobility, but also the way in which people interact with technology. Transparency offers an opportunity to make these new systems inclusive, safe and trustworthy.