Hackathon 2021

Tire maintenance is crucial in the mining industry, and productivity relies on tires that carry extremely heavy loads, over rough terrain, every day. As part go this project, students had to create a new business for Michelin Consulting & Services, combining existing data (and new data sources) and providing new and relevant KPI’s for tire maintenance management.

The winning team’s solution is a mobile-based application that provides the maintenance manager with all relevant information regarding a tire’s maintenance on one centralised platform. The key feature of the application is the tire’s health monitor, an indicator of the remaining useful life of a tire, which the team developed using machine learning. In addition, the students leveraged the data at hand by providing meaningful insights through dashboards and KPIs integrated within the application.

Michelin Track Connect (MTC) is an application that helps circuit drivers achieve their optimal driving performance. The goal of this project was for students to answer three key questions: In what ways can MTC be monetised? What should the growth strategy for MTC be? How can MTC monetise its data assets?

The winning team proposed three ideas. The first is to develop a dangerous driving risk alert, which could detect crashes and dangerous driving behaviour in real-time. The second idea is to sell insurance packages for race-day events through the MTC app by partnering with insurance partners. The third proposal involves providing insurance companies with data from MTC users to help them increase the accuracy of their risk assessment models.

The Truckfly application supports truckers throughout their journey. Amongst other features, it provides truckers with up-to-date information on where to refuel, eat and rest along their route. For this case, students had to think of ways to grow/pivot Truckfly’s business and generate stable revenues.

The winning team delivered a Competitive Analytics Dashboard where business owners and business managers can gain a competitive advantage given the exclusivity and granularity of Truckfly’s data assets. This Competitive Analytics Dashboard offers real-time data analytics about the industry and regional competitors, the trucker customer segments, and also, the main characteristics a business should have to maintain attractiveness and visibility in the market. All of this is possible by implementing statistical and machine learning algorithms, such as clustering, random forests, forecasting and market basket analysis.

There is increasing pressure to limit the online tracking of users. Thus, the advertising ecosystem is shifting towards contextual targeting. As part of this Criteo case, students were challenged to design a contextual targeting product to connect advertisers to publishers and reach users (using limited user data).

The winning team used keyword matching to categorise and assign advertisers with relevant publishers. It also drew data from multiple other sources, like Google Trends, to enhance the matching process.