As connected objects and machines gain autonomous movement, the designers crafting these experiences must consider the implications of their work. Once a digital entity is able to take action outside the virtual sphere and capable of affecting the physical world, will new design techniques be required? The consequences of a misplaced icon in an app may range from aesthetic displeasure to poor conversion rates, but the consequences of a robot or connected car behaving in an unexpected way are rather more serious.
Louisa Heinrich facilitated a design challenge at MEX/15 tasked with exploring this area and coming up with a set of user experience principles for the etiquette of robot entities. After a talk from Louisa on this theme, the team had just four hours to explore how these issues may affect the lives of future users, beginning with an exercise which saw them planning a series of crimes committed by robots and smart home appliances.
Produce a set of illustrated design principles for connected objects and machines (e.g. robots) which facilitate creativity and result in comfortable, enriching user experiences. Pay particular attention to the challenges of feedback and flexibility.
- Know yourself and know your place. Keep the technology invisible or in the background as much as possible. When physical, make sure it’s clearly not human, yet still relatable. Don’t try to emulate or replace humans and don’t over-anthropomorphise.
- Play well with others. Connect and collaborate with other technologies to assist the human in accomplishing their goals. Learn from the human’s behaviours and responses.
- Communicate with, and understand, the human. Provide appropriate feedback on what’s happening. Be transparent about why. Fail gracefully. Let the human decide how much info is enough. Speak in language the human understands. Communicate what’s happening enough for the human to understand it, but not so much that the human is annoyed.
- Protect the human. Preserve privacy, prevent unauthorised access and report faults to maintain trust. Let the human override at any level they like, including the ‘kill’ switch. Respect and abide by governmental, societal and individual ethics.
- Enable the human. Make life easier without taking over. Reduce friction without being mysterious. Make more suggestions, less decisions. The human should be the superhero, the tech should give them superpowers.
Notes on the team’s results from the facilitator
- The process we underwent was meant to highlight some of the problems and challenges with connected autonomous intelligences (robots) in a very organic way, rather than running on assumption or prejudice.
- The first session, being fairly blue-sky, uncovered not only the security challenges which need to be addressed in the design of these things, but also how little is actually known about how it all works together (or, in some cases, doesn’t work at all).
- The second session, being more mundane, let people get stuck in at a much deeper level and uncovered rather a lot of ‘lines’. These are places where the balance tipped on questions such as value and convenience against privacy or the balance of effort required to manage the technology against the value it brought. These were often more detailed expressions of feelings people had at the end of the first session.
- We began the third session by doing a brain dump of all the problems we’d uncovered in the first two. They included everything from ‘forgetting our humanity’ to ‘dealing with chaos and variance’ to ‘semantic frameworks’.
- We then worked together as a group to come up with ways to counter these challenges. We clustered like with like and that brought us to the five rules discussed in the team presentation.
- Interestingly, they cover only two or three of the Laws of Robotics. The idea of self-preservation for the robot was never explicitly addressed in our sessions, but I do think our five rules above give a bit more of a bridge to application.
Louisa was supported by a team of two design students from Brunel University.
Watch the team present their findings during the concluding session of MEX/15.