project
Satisfied truck driver, better society
Truck drivers appreciate their freedom and autonomy on the road, but automated trucks are safer and more sustainable - so better for society. Smart design can solve this value conflict.
Automated systems can contribute to goals like saving fuel in actual driving conditions, but drivers need to be willing to share control over their vehicle. This can be achieved by using persuasive technology. Drivers are found to be more inclined to share control over their vehicle if the persuasive system is coupled to a virtual driver similar to them. However, to prevent unethical manipulation that ignores the driver’s autonomy, certain guidelines need to be followed during design and implementation of the persuasive system.
(Semi)automatic forms of accelerating and braking may lead to fuel saving, but will not be successful - nor responsible - if no attention is paid to the interests of truck drivers. How may persuasive technology be used to optimise the distribution of control between driver and automated system?
A main outcome is that attention needs to be paid to the voluntariness and the autonomy of the driver. Responsible design and employment is a matter of finding the middle between ‘manipulation’ and ‘convincing’. The researchers propose three general ethical guidelines:
- “Persuasion should be based on prior (real or counterfactual) consent.” So truck drivers should be informed about the persuasive technology and asked if they agree to use it. The advantage is that they will be more likely to experience the interference of the technology as an attempt to convince rather than to manipulate them.
- “Ideally the aim of persuasion should be to end the persuasion.” For example, the gas-pedal could be equipped with a mechanical feedback mechanism that teaches the driver how to drive himself in a way that saves energy.
- “Persuasion should grant as much autonomy as possible to the user.” The truck driver should be given the opportunity to override the system. He could also be given the option to adjust system settings – for example choose between ‘strict coercion’ and ‘mild suggestion.’
Similarity cues
Another key finding: if the system’s display includes a ‘virtual driver’ similar to the human driver, the latter is much more willing to delegate control. Similarity can take different forms:
- in looks (an avatar that looks like the driver);
- in behavior (the avatar moving his head in the same way as the driver)
- in judgments (the system prioritizing driving goals – energy efficiency, speed, comfort and safety - in the same way as the driver).
If these ‘similarity cues’ are combined, the system is even more able to ‘persuade’ the driver to share control.
However, care should be taken to apply this finding in a responsible way. The underlying psychological mechanism of trusting people like you depends on automatic and subconscious cognitive processes. When two people interact they influence each other, resulting in better interaction. But when applied in persuasive technology this way of influencing the driver has the risk of leaning towards manipulation.
Unused potential
Mechanical engineers working on this project investigated the impact of different driver acceptance limits and gear shifting strategies on fuel saving. For that purpose they used a midsized passenger car (Nissan Primera) and took into account different driving conditions. Results of these studies were validated with truck drivers at the test facilities of valorisation partner DAF trucks. It affirmed that there is unused potential for fuel-saving.
Interdisciplinary
This project was able to contribute to responsible innovation because the case - potential fuel saving through automation in trucks - was studied in an interdisciplinary way:
- Psychologists looked at mechanisms of technologically mediated behavior change;
- Philosophers reflected on the ethical aspects of persuasive interventions;
- Engineers investigated the potential for gains in energy-efficiency.
This collaboration was one of the success factors of the project. It has strengthened and extended the network of the partners, and lead to a better understanding of the related discipline in each field.
What are ‘persuasive technologies’?
Characteristic for persuasive technologies is their interaction with the individual user; the technology’s suggestions and actions are adapted to the individual. This makes it very different from – and possibly more effective than - mass campaigns as a tool to get people to change their behavior. There are two different ways in which people become users of persuasive technology. With so-called ‘nomadic’ products the user chooses to adopt a persuasive technology. With ‘centralized’ products the user is involuntarily confronted with a persuasive technology, for example when a company installs it in a truck that its drivers have to use.
truck drivers, persuasive technology, automated systems, control, control, psychology, nudging, self-driving cars, autonomous driving, driver control, safety
Official project title: