SFU big data master’s students win big at inaugural Unearthed Vancouver hackathon

October 20, 2017
Print

By Katie Knorr

A team of three SFU big data master’s students won first prize at Unearthed Vancouver, a 54-hour open innovation hackathon focused on the mining and resources sector. A second team of big data master’s students, WonderWomen, won the popular People’s Choice Award.

Held in Vancouver for the first time, the hackathon attracted over 100 software developers, engineers, designers and data enthusiasts who were eager to spend a weekend solving problems unique to the mining and resources sector. Industry insiders and subject matter experts from mining companies SSR Mining and Teck Resources guided participants in crafting solutions for real-world mining operations.

The winning trio, consisting of Aakash Moghariya, Steve Matson and Ruturaj Patel, took home a $2,500 cash prize and $3,000 in Amazon Web Services credits.

Team Data Smashers from left to right: Ruturaj Patel, Steve Matson and Aakash Moghariya.

“We heard about the hackathon from our program director, professor Mohamed Hefeeda,” Patel says. “We really wanted to apply our data science skills to a problem from such a traditional industry with huge potential for data-driven decisions.”

The students’ solution addresses the costly problem of unplanned equipment downtime, the time a piece of equipment is taken out of service for unplanned maintenance work. Using regression and classification techniques, Moghariya, Matson and Patel developed a machine learning platform that assists operations supervisors and maintenance technicians in predicting haul truck failure and planning for maintenance.

Team WonderWomen with Chris Schmid, Open Innovation Lead at Unearthed Solutions. Team members from left to right: Leiling Tao, Katrina Ni, Yuyi Zhou and Ruoting Liang.

Team WonderWomen also focused on the truck failure problem. Leiling Tao, Katrina Ni, Ruoting Liang and Yuyi Zhou developed a predictive model that boasts over 20 features and can forecast truck failure with 79.5 per cent accuracy. In addition, they built an easy-to-use web-based user interface which helped the team stand out from the competition.

“We had a very clear and well-executed concept, and we delivered it well through our presentation,” Liang says. “Although we did not win the first or second prize, our product came up very strong overall.”

Both teams are collaborating with SSR Mining to further develop their respective prototypes in the upcoming months and explore possibilities for their implementation.

Moghariya and Patel also recently won the second place prize ($2000) at the POF Hackathon, which was hosted by SFU’s Faculty of Applied Sciences in partnership with Plenty of Fish. The pair shares this prize with team members Sreenath Vazhayil and Karthikeyan Chandrasekaran who are also big data master’s students.