aSTROBEE
Astrobee is a robot that is flying around the International Space Station (ISS) alongside the astronauts.
In addition to providing a research platform for zero-g free-flying robotics, Astrobee improves the efficiency of ISS operations by providing flight and payload controllers with a mobile camera/sensor platform.
Role : Human-Robot Interaction Lead
Mainly contributed in user study, UI design, and prototyping.
Also worked on product and visual identity design.
Period : 2016 - 2018
DESIGN GOAL
As the first autonomous free flyer in ISS,
make Astrobee to cohabit with crews without distracting crews from their work.
DESIGN PROCESS
RESEARCH
Not like everyday product, Astrobee has variety of internal states from on / off to moving, docking, and requiring help because of its high level autonomy and movement. Also, time of crews in ISS is so valuable, the signal from Astrobee should make people aware, but not disrupt them.
To find a way to design signals for Astrobee with the purpose described above, I set two research questions and conducted experiments for those questions: (A) Is the person aware of the signal representing an internal robot state? and (B) what states do the person think the robot is in through the given signal?
Key Research Goals
Identify the basic robot states to be conveyed to astronauts for human safety and task efficiency
Explore the way of making astronauts aware of robot states without distracting
Suggest logical design decision to represent robot states
Literature Review
To identify information that is potentially important for non-humanoid robots to signal, I reviewed usage of nonverbal signals of non-humanoid robot state and behavior. The non-humanoid robot includes industrial mobile robots, autonomous vehicles, flying robots, service robots, and tele-presence robots. From the literature review, I organized 14 robot states into five categories based on its caution and warning level.
Light Attributes
Light is the most popular modality used for conveying a robot’s state. I selected light as a channel for signaling Astrobee's states, then defined attributes of light that designers can adjust: dimensional and mannerly attributes.
Dimensional attributes: visual properties to change
Mannerly attributes: dynamic way of changing dimensional attributes
Case Study
To deduce the design principles for creating light signals to indicate a robot’s internal state, we analyzed the three existing robots having different physical form (two-wheel based vs. pillar-shaped vs, pole-shaped) and shape of light array (wheel vs. ring vs. bar).
From the case study and literature review, I formulated hypothesis about the effect of light attrobites on signal characteristics.
Participatory Design for Interpretation
To create light signals intuitively conveying robot states, I performed a participatory design for a collaborative approach with design experts. Four UX designer having human-robot interaction design experience were participated. The participants were asked to create light signals for each given 12 robot states collected from the literature review except 2 robot states overlapped with other state or having obvious light patterns. The Interactive light pattern design tool was provided for the participants to adjust light patterns and light attributes including color, tempo, and continuity. The participatns were remotely joined the design session, and used their personal laptop to design light signals.
Survey for Interpretation
To validate the light signal design from participatory design and define design details for each robot state, survey was conducted with 31 participants (17 males, 13 females, and one who did not want to select gender; ranged between 25 and 51, M=32.37, SD=5.52). Participants were asked to select how appropriate the given light signal is to convey each robot state. Total 43 light signals from the participatory design were programmed with Arduino and LED arrays on the fabricated a box-shaped robot. Each light signal was video-recorded and provided via Google survey.
Internal Robot States and Awareness Level
Internal states of Astrobee have different level of urgency and importance to be conveyed. Therefore each state requires difference in awareness level. We defined the required awareness level for the most common internal robot states.
User Study
We conducted a user study aiming at gaining better understanding of how mannerly attributes of light signals affect people’s awareness and how it can be used for noticing robot’s internal states without disturbing people’s primary task performance.
The overall experimental procedure and results will be published soon.
HUMAN-ROBOT INTERACTION DESIGN
Major light attributes which affect people's interpretation are movement, color, and position of light. So logic of light movement, color, and position for light signal pattern was defined. Based on the logic of light signal, the light signals indicating basic states of Astrobee is simulated by Javascript. Overall guide for engineers were provided with 2D wireframes and recorded videos of rapid prototype using Arduino.
PRODUCT DESIGN
To let Astrobee cohabit with astronauts in the International Space Station, there were some issues we needed to solve in the perspective of product design.
Product Design Requirements
Sleek, not bulky
Astrobee is 12" cube shape, so it can be looked bulky. It should be sleek with CMF.
Functional affordance
It is important to figure out where Astrobee is heading. Orientation should be clear.
Identification
Multiple Astrobee will be working, so clear identification of those robots are needed.
Product Design Outcomes
Astrobee has an arrow graphic on its side to indicate where it is heading as well as looking it slim. The side panel of Astrobee is made of a changeable fabric. Guest scientists are allowed to print their logo, emblem, and any graphic they want on the fabric in order to identify their own robot. The printed AR markers on the fabric make crews detect its identification as well.
VISUAL IDENTITY DESIGN
Astrobee is a free flyer for IVA (Intra Vehicular Activity). In a mission patch, I emphasized that an astronaut is in space and Astrobee in ISS, but collaborate with each other.
RELATED PUBLICATIONS
Cha, E., Kim, Y., Fong, T., & Matarić J. M. (2018). A Survey of Nonverbal Signaling Methods for Non-Humanoid Robots, Foundations and Trends in Robotics, 6(4), 211-323.
Cha, E., Fitter, N. T., Kim, Y., Fong, T., & Matarić, M. (2018). Generating Expressive Light Signals for Appearance-Constrained Robots, In: Proceedings of International Symposium on Experimental Robotics (ISER 2018), Buenos Aires, Argentina.
Cha, E., Kim, Y., Fong, T., & Matarić, M. (2018). Effects of Robot Sound on Auditory Localization in Human-Robot Collaboration. In: Proceedings of 13th ACM/IEEE Human-Robot Interaction Conference (HRI 2018), Chicago, U.S.A, 434-442 (acceptance rate: 23%).
Kim, Y., & Fong, T. (2017). Signaling Robot State with Light Attributes. In: Proceedings of 12th ACM/IEEE Human-Robot Interaction Conference (HRI 2017), Vienna, Austria, 163-164.