How to Design Trustworthy Robots
Trustworthy robots have the potential to change our lives, but are not perfect and still have limitations.
Mathematical tools and algorithms
Trustworthy robots require mathematical tools and algorithms, such as logic, formal synthesis, and control barrier functions, to ensure safety.
Alignment of engineering and human perspective
Both the engineering perspective and human perspective on trustworthiness need to be considered and aligned when designing robots.
Understanding human behavior
The human perspective helps in designing robots that people trust to the extent they should be trusted.
Logic gives me a way to describe what I want from a robot… The beauty of formal synthesis is that if I decide that my robot Butler should be doing something else, I just change the logical description… I don’t need to reprogram the robot. – Jana Tumova
Factors influencing trust
People’s behavior around robots is influenced by factors such as appearance, agency, and motivation.
Logic and formal synthesis
Logic and formal synthesis are used to describe what we want from a robot and turn it into a correct plan of action.
Robot appearance
Human-like appearance may lead to higher expectations of human-like behavior, but it does not necessarily make the robot appear safer.
Combining mathematics and social sciences
To design trustworthy robots, a combination of rigorous mathematical tools and social sciences is required.
Control barrier functions
Control barrier functions act as an invisible fence to prevent collisions and protect the robot from harm.
Perception and trust
Trustworthiness of robots is not solely determined by engineering guarantees, but also by how people perceive and trust them.
Comments ()