Lorentz Center - HART - Human-Agent-Robot Teamwork: Tools and Methods for Designers from 5 Jan 2015 through 9 Jan 2015
  Current Workshop  |   Overview   Back  |   Home   |   Search   |     

    HART - Human-Agent-Robot Teamwork: Tools and Methods for Designers
    from 5 Jan 2015 through 9 Jan 2015


Scientific report

The Lorentz Workshop on Human Agent Robot Teamwork of January 2015 encountered some set backs. Most of our international keynotes speakers had the flu. They did arrive in Leiden, some could even give their presentation, but then had to lay low in their hotel rooms waiting for the flu to run its course. Luckily, none of them missed their return flights.

Content-wise the workshop was still a success. We had great discussions on methods, inititated by the presentations by Mark Goodrich, Frank Dignum, and Matt Johnson. The debate was so intense and constructive that we replanned our workshop to be able to continue with more plenary sessions talking more about methods and tools. Virginia Dignum spoke about OperA, and Mark Goodrich said more about his sCET, Jurriaan van Diggelen gave a great demonstration and talk about his tools for workflows and planning for HART.

Catholijn Jonker made a strong case for design for values and argued that design-time we should design for run-time solutions for value conflicts. She also argued that most designers think that they design the whole system, whereas with the advance of the Internet of Things, many teams will design for the same or almost same problem or application, but not at the same time, and not being aware of the other teams. The brings the problem of how to design for open systems in a value-sensitive way to the fore.

There was a great discussion about how all these tools and methods could fit together in a complementary way, and the hope was to continue that discussion of relationships among tools (as well as gaps) even after the workshop.

The impact of the workshop has been that within the Netherlands, in the collaborations for setting up Zwaartekracht and Perspective proposals the roboticists are in agreement that we should shift our focus from fully autonomous systems to robots that are designed for co-activity, and for sharing control. Really fascinating connections grew out of these concepts, including our overall insight that the three levels of expertise are closely linked to shared (haptic) control, co-activity, and having the flow of control go to the team member that has the best expertise for the current task of the team work. Questions of how to deal with responsibility and accountability will be returning topics of concern and research.