In ice hockey’s earlier days, National Hockey League (NHL) coaches made their most important decisions based on gut instincts. Today, experience and instincts are still vital, but NHL coaches now have another essential tool at their disposal: powerful data analytics. Before and after every game, coaches and even players meticulously pore over game data and review detailed statistics to improve performance and strategy. And while this is a win for the NHL, higher-end data analytics tools have typically been out of reach for youth hockey teams largely because capturing game performance data on the ice is expensive, complicated, and time consuming.
We built Drive Hockey Analytics to democratize pro-level analytics and help young players develop their gameplay and build a higher hockey IQ. Coaches and parents can now easily and affordably track 3,000 data points per second from players, sticks, and pucks. Drive Hockey Analytics—which takes 15 minutes to set up at the rink after initial calibration—converts these raw data points into actionable statistics and insights to improve player performance in real time and boost post-game training.
Scaling a market-ready stick and puck tracking platform on Google Cloud
Drive Hockey Analytics began as an engineering project in the MAKE+ prototype lab of the British Columbia Institute of Technology (BCIT). We quickly realized that we couldn’t transform Drive Hockey Analytics into a market-ready stick and puck tracking platform without shifting more resources to R&D. After meeting with the dedicated Google Startup Success Managers from the Google for Startups Cloud Program, with this support, we decided to migrate from AWS to Google Cloud so our small team could reduce IT costs and accelerate time to market.
Google Cloud solutions make everything easier to build, scale, and secure. We immediately took advantage of Google Cloud’s highly secure-by-design infrastructure to implement robust user authentication and institute strict privacy controls to comply with the Children’s Online Privacy Protection Act (COPPA). In just days, we enabled coaches and players to access individual analytics dashboards and more securely share key statistics such as speed, acceleration, agility and edgework, zone time, positioning, among many others with teammates and family.
We also separated performance and personal storage data on Google Cloud, encrypted containers with Google Kubernetes Engine (GKE), and wrote third-party applications and pipelines that autoscale with Spark on Google Cloud. These processes could have taken us weeks or even months if we had to manually design and integrate all these security capabilities on our own.
To build our interactive player analytics engine, we leveraged TensorFlow, BigQuery, and MongoDB Atlas on Google Cloud. With the simple and flexible architecture offered in Google Cloud, we quickly moved from concept to code, and from code to state-of-the-art predictive models. We now collect and analyze thousands of data points every second to identify key performance metrics, break-out game intelligence, and deliver actionable recommendations. Coaches and players can leverage this data to increase team possession of the puck, optimize player positions, reduce shot attempts, and score more goals.
In the future, we plan to explore additional Google products and services such as Google Cloud Tensor Processing Units (TPUs), Google Cloud Endpoints for OpenAPI, and Google Ads. These solutions will enable us to further expand our ML stack, leverage streaming data from wearables and cameras, and reach new markets.
Bringing pro-level sports analytics to youth hockey
The Startup Success team has been instrumental in helping us rapidly transform Drive Hockey Analytics from a university engineering project into a top shelf player and puck tracking system. Their guidance and responsiveness are amazing, with a human touch that stands out compared to services from other technology providers.
We especially want to highlight the Google Cloud research credits that help us affordably explore new solutions to address extremely large dataset challenges. Thanks to these credits, we successfully process thousands of data points in streams and batches, apply ML-driven logic, and run resource-efficient queries. Google Cloud research credits also give us access to dedicated startup experts, managed compute power, vast amounts of secure storage, and potential for joining the Google Cloud Marketplace.
Demand for Drive Hockey Analytics continues to grow, and we constantly evolve our platform based on input from youth teams and coaches. We’re looking to go fully to market in 2023. With Drive Hockey Analytics, youth teams are putting on their mitts and taking control of the puck as they improve real-time player performance and help their team count more wins. We can’t wait to see what we accomplish next as we continue transforming dusters into barnburners by democratizing advanced analytics that were once only available to pro-sports teams.
If you want to learn more about how Google Cloud can help your startup, visit our page here to get more information about our program, and sign up for our communications to get a look at our community activities, digital events, special offers, and more.
By: Adam Nathwani (COO, Drive Hockey Analytics) and Mike Dahlstedt (Founder and CEO, Drive Hockey Analytics)
Source: Google Cloud Blog