The open roboethics initiative and the elevator-riding robot

A'Jung Moon, Ergun Calisgan, Camilla Bassani, Fausto Ferreira, Fiorella Operto, Gianmarco Veruggio, Elizabeth A. Croft, H. F.Machiel Van der Loos

Research output: Chapter in Book/Report/Conference proceedingChapter (Book)Researchpeer-review

1 Citation (Scopus)

Abstract

The increased development and deployment of robotics has brought with it a growing expression of ethical, legal and societal implication (ELSI) concerns among designers and the public. There is a need to break down current boundaries in roboethics discussions and to broaden stakeholder dialogues on these issues. The “Open Roboethics initiative” (ORi) is an open-source, Internet-based resource to drive discussion of roboethics, policy, and design. The authors envision ORi to grow into a dynamic online platform where various stakeholders can connect and engage. This synergy between bottom-up roboethics discussions and open sharing of robot designs is proposed to accelerate policy and design changes in robotics. Bottom-up approaches are among the most inclusive and useful in addressing applied ethics issues as social norms evolve. They aim to understand the opinions and perceived values of the public or stakeholder groups, while being managed and supported by an “experts committee.” Open source models encourage developers to share designs or source code and to distribute their technical contribution to the worldwide community for free. Established open source initiatives are discussed, showing examples of success and providing lessons on the practical issues of an open, online community for advancing roboethics. To demonstrate how ORi works, the authors present a case study involving a humanoid robot, PR2, who needs to ride an elevator to fulfill a delivery task, putting the robot in conflict with existing elevator patrons. In this case, ORi uses an online discussion space to crowd-source the perceptions of participants about cultural norms, expectations, social conventions and ELSI issues. The modules uses Q-learning, a machine learning technique, to select an appropriate behavior based on each unique situation, given pilot online survey responses. This case study demonstrates how the ORi concept can be particularly effective and illustrates some challenges that may occur in its implementation. The authors believe ORi will serve as a catalyst for discussions within and across nations and organizations regarding robot technology design and policy changes.
Original languageEnglish
Title of host publicationRobot Law
PublisherEdward Elgar Publishing
Chapter6
Pages131-162
Number of pages32
ISBN (Electronic)9781783476732
ISBN (Print)9781783476725
Publication statusPublished - 1 Jan 2016
Externally publishedYes

Cite this