PolarBearings
Mobilizing AI to navigate the Arctic
Problem
Finding safety in a melting Arctic
Climate change has severely altered the Canadian Arctic landscape, making travel routes unreliable. Rapidly changing ice conditions threaten locals’ way of life. Maritime navigation tools exist, but a void remains for land-based transportation.
How might folks travel safely in this unpredictable environment?
Stats
50%
Temp. increase over 100 yrs.
13%
Sea ice decrease since 1950
60%
Salmon population decrease
My Role
Build an app in 2 days
As the only UXer at the 2nd Cassini Hackathon in Berlin, I joined a global team of developers, ML engineers, and business consultants to tackle the theme of ‘connecting the Artic’ via AI and satellite data from Copernicus and Galileo.
I set out to design a prototype for researchers, logistical workers, and most importantly, indigenous communities often left out of tech solutions.
Tasks
Conduct research
Design prototype
Pitch MVP to judges
Discovery
Uncharted AI territory
During my research, I found two similar platforms, but they had limited reach and relied on volunteer-submitted data, which was inconsistent and outdated. None of them harnessed AI to interpret satellite imagery or offered alternative routes in response to shifting ice conditions caused by climate change.
I interviewed Arctic experts at the venue to learn about the region’s transport issues. They highlighted sparse travel options and unstable ice roads (routes for moving cargo). These insights reinforced the need for real-time AI-powered navigation to help travelers adapt to unpredictable terrain.
Insights
~200,000
Canadian Arctic population
54 days/yr.
Average ice road usage
Radio over wifi
Spotty internet coverage
Testing
From satellites to snowmobiles
I used the insights to design an MVP, but geography and a ticking clock prevented me from conducting test sessions with Arctic locals. Instead, I opted for a faster alternative: guerrilla tests with folks at the hackathon venue. I focused on making sure the mobile experience was minimal and intuitive with a low learning curve.
The app allowed users to download up-to-date satellite imagery before setting out, allowing them to travel over rough terrain even without connectivity. The AI-powered navigation adapted to changing ice conditions, offering alternative routes in real time based on the latest satellite data.
Results
5/5
Questioned map icons
3/5
Confused by CTA
4/5
Found transport options clear
Final design
Meet PolarBearings
The MVP highlighted quick decision-making for arctic travelers in harsh and unpredictable environments, emphasizing clarity, efficiency, and real-world data.
Despite a looming deadline, we submitted our pitch just in time, securing 1st place at the hackathon and 2nd in the European finals.
Impacts
$1,000
Startup incubation fund
Mentorship
Awarded expert guidance
Opportunity
Pitched to SAP SE in Berlin
Reflection
Lessons learned
In just two days, I learned how to utilize on-site resources and maximize time during meetings. I also learned how to work with developers and engineers, since getting these folks on board early in the design process is critical.
This was also the first time I experienced hybrid meetings. Relaying messages back and forth in-person to online was odd and challenging. I prefer either all remote or all in-person meetings to maximize communication and time.
With a hackathon win under our belt, PolarBearings was ready for a new expedition.
Let's explore and create together
Made with a compass and a pot of Earl Grey by Ian Wright, 2025
