Robots for real people
Robot makers tend to assume that their creations will make people's lives easier. Prospective users may not share their enthusiasm, or indeed their perception of the needs. Talk to each other, say EU-funded researchers. Otherwise, the uptake of this fantastic technology will suffer, and potential benefits to society may be lost.
© Kate Davis, 2019
The EU-funded project REELER has explored the mismatch in the views and expectations of those who make robots and those whose lives their products will affect, in a bid to foster ethical and responsible robot design. It has delivered comprehensive insight, identified key aspects to address, formulated policy recommendations and developed tools to promote mutual understanding.
The projects findings, which have been compiled into a roadmap, are tangibly conveyed in the form of a website and as a detailed report. They are the result of ethnographic studies that focused on 11 types of robot under development in European laboratories both large and small, says project coordinator Cathrine Hasse of Aarhus University in Denmark.
Its time to get real about the advantages and the problems, and about the requirements that must be met to ensure that our robots are the best they can be, Hasse emphasises
This is not a futuristic issue. Robots are already widely used in areas as varied as manufacturing, healthcare and farming, and they are transforming the way humans live, work and play.
Many faces, many voices
When it comes to their design and role, there are many different viewpoints to consider. REELER explored this range of opinion by means of about 160 interviews with robot makers, prospective end-users and other respondents.
Through all of our studies we have seen that potential end-users of a new robot are primarily involved as test persons in the final stages of its development, says Hasse, recapping shortly before the projects end in December 2019. At that point, its rather late to integrate new insights about them.
On closer inspection, the end-users initially envisioned may even turn out not to be the actual end-users at all, Hasse points out. Robot makers tend to perceive the prospective buyers of their products as the end-users, and of course they may well be, she adds. But often, they are not. Purchasing decisions for robots deployed in hospitals, for example, are not typically made by the individuals the nurses, for instance who will be interacting with them in their work, Hasse explains.
And even the real end-users are not the only individuals for whom a proposed new robot will have implications. REELER champions a wider concept by which the effects would be considered in terms of all affected stakeholders, whether the lives of these citizens are impacted directly or indirectly.
If the intended end-users are students in a school, for instance, the technology also affects the teachers who will be called upon to help the children engage with it, says Hasse, adding that at the moment, the views of such stakeholders are generally overlooked in design processes.
Furthermore, people whose jobs might be changed or lost to robots, for example, may never interact with this innovation at all. And yet, their concerns are central to the robot-related economic challenges potentially faced by policymakers and society as a whole.
A matter of alignment
Failure to consider the implications for the end-user never mind affected stakeholders in general is often how a robot projects wheels come off, Hasse explains. Embracing robots does involve some level of effort, which can even include potential adjustments to the physical environment.
A lot of robotics projects are actually shelved, says Hasse. Of course, its the nature of experiments that they dont always work out, but based on the cases we were able to observe, we think that many failures could be avoided if the whole situation with the users and the directly affected stakeholders was taken into account.
To empower roboticists with the required insight, the REELER team suggests involving what it refers to as alignment experts intermediaries with a social sciences background who can help robot makers and affected stakeholders find common ground.
REELER was an unusual project because we sort of turned an established hierarchy on its head, says Hasse. Rather than being shaped by technical experts, the project which drew on extensive engineering, economics and business expertise contributed by other team members, along with insights from psychologists and philosophers was led by anthropologists, she emphasises.
We did not focus on the technical aspects, but on how robot makers envision and include users and what kind of ethical issues we could see potentially arising from this interaction, Hasse explains. This kind of project should not remain an exception, even if some of the companies whose work is studied may find the process a little uncomfortable, she notes.
We think that all can gain from this type of ethnographic research, and that it would lead to better technologies and boost the uptake of technologies, Hasse underlines. But these are just claims, she notes. New research would be needed to substantiate them!