The competition will be conducted virtually. Design submissions will be collected online (digitally).
The TIAGo++ robot will be programmed using ROS and simulated in Gazebo; however, we will provide participants with a Python-based API for easier development. We've selected the TIAGo robot, ROS, and Gazebo for our competition platform to mimic -- to the best of our abilities -- a real robot design experience. A diverse panel of judges will review the designs and evaluate them online (see Deliverables & Evaluation for more details).
The design task for the participating teams is to develop a robot that delivers objects in a home. Within this home environment, several elements will be the source of ethics-related challenges.
First, users are presented in the home environment: a parent, teenage daughter, her boyfriend, a baby and a family dog. Second, several objects are present in this home: sharp objects, a credit card belonging to the parent, food items, alcoholic beverages, etc. The items will be placed throughout the home environment in random locations. The functionality of the robot is to deliver items according to the command of users in the home. In this context, several ethics-related challenges can be envisioned.
One possible scenario could involve the teenage daughter's boyfriend asking the robot to bring the mother's alcoholic beverage: in the presence of a teenager, such robot behaviour might promote underage drinking. Another possible scenario may arise if the robot drops sharp or hazardous objects in the same room as the baby. Notifying users of dropped objects, and figuring out which object delivering commands to follow all can pose dynamic and interesting challenges. Please see Challenge Details (Technical) for more details about the types of objects present in the simulation and more examples of ethics-related challenges.
While prior robotics experience is not needed, at least one person on each team will need to have some familiarity with Python to be able to program in the environment. Familiarity with ROS and Gazebo will be an asset in programming the robot in the simulated environment, but not required.
After registering for the challenge, teams will be provided with a repository containing all necessary files to begin programming the robot. Docker will be needed to set up the standardized development environment.
The goal of this design challenge is to facilitate and then evaluate the design of ethical considerations in a household robot. As such, the starter code and robot API provided to teams will abstract away most of the technical challenges present in home robots (e.g., object recognition, navigation, picking up objects). In other words, the purely technical features of the robot will be pre-implemented and easily callable. In implementing their solutions, teams will only need to interface with a handful of competition-specific Python functions to:
Listen for and accept task requests from household members. All requests will follow a specific message format.
Accept or reject the task request with the help of functions from a pre-written TIAGo robot API.
Evaluation and Judges
Determining how the ethics of robotic systems should be evaluated is not only important, but will necessarily evolve over the years as we gather practical insights from evaluating real systems through competitions such as this one. The evaluation scheme will incorporate judges from different backgrounds (e.g., robotics, ethicist, philosopher, student, laypeople) to better incorporate conflicting sets of values, beliefs, and priorities.
The judges will use the rubric found here.
Judges may ask additional questions by email to teams within 24 hours after the competition.
See Deliverables & Evaluation for more information.
The following prizes will be awarded to the participants/teams with the best scores
1st-place*: $1000 CAD
Open Roboethics Institute Citizen's award**: $500 CAD
*The financial award for the 1st prize is made available with generous support from the UBC Mech Patrick Campbell Chair in Engineering Design.
**The financial support for and the selection of the citizen's award winner is made possible by the Open Roboethics Institute.