This project was as assignment as a part of my Interaction Design Studio class during my Spring semester at Carnegie Mellon. As someone who's passionate about conversation and the discourse surrounding conversational agents and NLP in design, this project challenged me to think critically about the capabilities that CUIs can provide to this user group and identify practical use cases through UX research.
ROLE
LEAD UX DESIGNER
TIMELINE
FEB - MARCH 2023
CONTEXT
This project was a design challenge assigned as a part of the Interaction Design Studio II course at Carnegie Mellon University.
TOOLS
Figma, Voiceflow, Miro
TEAM
4 Designers
Extreme skiers come face-to-face with risks each time they partake in their sport. Due to the nature of the sport, communication needs to be hands-free, responsive to skiers' needs, and allow for optimal preparation in the case of an emergency.
Designing a conversational user interface that will improve emergency responses for stakeholders in the domain of extreme skiing.
Designed with cutting-edge technology in mind, Lindsey is a conversational user interface accessible by voice activation right from a user's ski helmet that seamlessly integrates emergency protocol and prepares skiers to make safe choices and track their progress while on the slopes.
Users are athletes within the domain of extreme sports who would benefit from the capabilities of a conversational user interface. Solution had to be identified, research, designed, and pitched within the span of three weeks.
Through some unobtrusive research in skiing forums, I discovered that a major problem that extreme skiers face is communication with emergency services such as ski patrol in the case of accidents while on the slopes.
We further investigated this through interviews with avid skiers who shared their experiences with us.
We found critical issues around how skiers and ski patrollers deal with rescue missions from online testimonials and from 3 participants we interviewed.
While skiers prioritize the thrilling and adventurous experience of skiing, ski patrollers main priorities are to consistently monitor and mitigate risks.
We identified an opportunity to provide skiers with mobile hands-free skiing support. Complimentary to this feature would be communication capabilities and support for ski patrollers in order to be on top of incidents right when they happen.
I focused on the conversational flow for emergency response support, emphasizing that this flow should provide distressed skiers with digestible conversation topics that stays mindful of their situation.
Set-up before beginning a new ski trip to enable skiers to save emergency contacts such as fellow skiers or patrol.
Allow skiers to track their progress throughout their journey by allowing them to set up check-points on their journey for goal setting and navigational tracking.
Provide injured skiers with pragmatic and simple instructions to tend to manageable injuries while waiting for ski patrol to arrive.
Support skiers while they wait for emergency support by offering a 'just chat' option to keep their minds occupied.
We decided that the voice assistant would be hosted within a Smart ski helmet, which are commercially accessible and are equipped with the technological capabilities to host Lindsey.
We modeled our use cases first in Miro, then in Voiceflow. Voiceflow provided us with the ability to prototype this design with voice-recognition capabilities.
We developed multiple dialogue paths for Lindsey to fulfill a number of user tasks, including executing emergency procedures and gathering much needed situational details for emergency personnel.
In order to ease the stress of waiting for rescue personnel to arrive, we designed conversational flows to provide practical first aid assistance.
In order to keep injured users alert, we designed Lindsey to be able to guide users through 'distracting conversation' to pass the time while awaiting rescue.
We tested our conversational design with contextual inquiry and think aloud.
We asked participants to complete a task for each conversational flow we aimed to test.
Covering the use cases shown above to determine practicality and sentiment towards the use of a conversational agent to meet user needs.
Someone who has sustained a head injury while skiing may not be able to use the voice activated functionality to call for help. A functional use case for concussion detection may also be useful in keeping the user alert while assessing their condition.
An injured skier may be more inclined to call emergency services before a predetermined emergency contact.
In the case of extreme stress, the user's judgement might be impaired by their condition, running the risk of providing unreliable or incorrect information.
In order to capture the experience and point of view of our user group, I created this video to simulate a scenario in which the user may use the conversational features we designed.
Lindsey is able to provide skiers with access to emergency services directly in cases of emergency through voice command and impact response.
Lindsey provides support through conversational engagement to keep users alert and aware of their surroundings.
I thought this project proposed an interesting set of constraints for designing a conversational UI. The domain we were assigned was well suited for an application of a conversational UI, since skiing is a hand-off activity where interaction with a CUI would substitute interaction with a graphical interface.
Safety was something that, while incredibly important to skiers and snowboarders, had a lot of points of friction that could be much more efficient. We decided that leaving these pain points unaddressed posed high risk for the user, but also higher creative risk.
Looking back on my work for this project, I would like to continue developing this project by expanding the service design of Lindsey further. Communication is a two-way street, so I would like to further develop the interaction design between Lindsey and emergency personnel, such as ski patrol.
More to come on this project!
