Date:
Jan 2022 - April 2022
Role:
Product Designer & UX Researcher. UX Research, User interview, User testing and designed 3D printed prototype via AutoDesk Maya
Teams:
Our robot and app design Swann aims to enhance professional meeting environment through balanced participation, and better collaboration.
Imaginean employee or student using video chat platforms such as, Zoom, Google Meet and Discord, who needs more effective communication methods such as a physical conversation augmentation to help facilitate hybrid conversation with his colleagues at work.
Our product, Swann, will let users have effective communication and speaking etiquette during meetings by encouraging users to be more proactive in ways like engaging more or less and sticking within an allotted participation timeframe. We will measure effectiveness by speaker engagement profiles and calculating speak duration through speech recognition and computer vision algorithms.
Research has shown that greater parity of conversational turn taking is predictive of group performance and this made us thought about design augmentations in the physical world to improve performance.
The stakeholders responded generally positively to our findings and design. Direct stakeholders such as users and executive director expressed interest to further develop the technical aspect such as the robotic movement and ML driven voice recognition. Since this was a student project, as the class ended unfortunately we were unable to implement the designs into production.
The team of three UX researchers including myself conducted literature review on the design of our robot through previous research. Our design choices for the Swann Robot were informed by prior research "Learning Compact 3D Models"
The design choices for the Swann Robot were informed by prior research "All Robots Are Not Created Equal" by DiSalvo et al.,
Here is some of the take aways from the research done by DiSalvo et al.,
Thus we decided to use these guidelines to design our Swann Robot physical platform
Initial User Interview Demographics:
User Interview Findings:
After learning from the research insights I conducted. These design requirement were created to resolute the problems we see in our users.We first developed a series of low-fidelity prototypes in paper and InVision to visualize our design and conduct user-testing
By collaborating with two other UX designers on the team, we brainstormed together and sketched out the wireframe via paper prototype. Then we created the high-fidelity prototype using InVision and FIgma.
Low Fidelity User Testing Demographics:
User Testing and Interview on Low-Fidelity Prototype
Design Revisions for the High-Fidelity Prototype
The robot would move slowly towards the speaker to signify that someone elsewhere name appears on Swann screen is willing to talk.
Swann can be integrated into video communication platforms like Zoom through its plugin.
Swann application has facial and voice recognition which users can save their information and it uses Unsupervised ML algorithms which utilizes noise canceling to disregard unwanted sounds or facial features entering the meeting.
"The design of keeping conversation balanced is very relatable to our daily professional lives." - Linkedin Project Manager
"The robot is a cool idea but I wonder if there are other fields of application for this design For example a presidential debate robot?" - PhD Student
"The speaker who is cut off by the robot could have negative feelings in the moment, could there be a emotional support design that comforts others after they are cut off?" - Engineering Professor
The stakeholders such as responded positively to our findings and recommendations. Many of the users we interviewed expressed hopes for this design to be produced in the future.