ML & AI / Case study

AI & Machine Learning for wellbeing products

About the game

  • 5 trials in schools including a randomised controlled trial
  • Successful crowdfunding campaign
  • 1200+ early users
  • 4 in 5 trial participants able to use techniques taught by the game to reduce physiological signs of stress
  • 3 in 4 reported getting better at staying calm and focused within the game

BfB Labs create products and services for better healthcare and wellbeing. They decided to focus on mental health in teenagers by using an innovative sensor to improve mental wellbeing through gaming.

BfB Labs created a video game, ‘Champions of the Shengha’ that helps players focus their mind and improve emotional self-regulation skills. The game tracks a players’ stress levels using heart rate variation (HRV) sensors and rewards those who can stay calm under pressure.

Dae.mn helped build an analytics engine that provides insight into HRV performance over time and an ML model to personalise player development (e.g. identifying the right types of opponents) and improve engagement.

The challenge

Champions of the Shengha collects second-by-second data on players’ heart rates, processed through an algorithm to derive the players’ Heart Rate Variability (HRV). BfB Labs Technology enables tracking of player improvement. Our client required a standardised, automatic way of analysing a player’s HRV over time. As a result, BfB Labs had a lot of unprocessed data that could add value to the business and player.

The scope included:

What we did

BfB possessed a huge amount of engagement data that could be used to improve future iterations. We worked with them to build their understanding of machine learning and the value it can add to their business & customers. The process began with a focus on understanding BfB Labs data in depth, and finding how machine learning could best benefit their business. Dae.mn used AWS technology to build dashboards that captured insights into player engagement.

Using the vast amount of user data (entries, performance, interaction, engagement) and learning metrics (sensor data, sensor graphs and metric trends) we built an algorithm that converts the heart rate into a metric showing improvement that enables Champions of the Shengha to personalise player development further. We built machine learning models that would predict how a player improves based on data, facilitating adaptive gameplay. This would enable adaptation of levelling and AI opponent performance, optimising the game to engage players further. Our solution included a machine learning module designed to create insights for education and healthcare professionals in predicting user development.

Outcome

Using this algorithm we presented insights into player improvement, creating opportunities for Champions of the Shengha to personalise player development further. We built machine learning models that would predict how a player improves based on data, facilitating adaptive gameplay. This would enable adaptation of levelling and AI opponent performance, optimising the game to engage players further. Our solution included a machine learning module designed to create insights for education and healthcare professionals in predicting user development.

Our solution enabled BfB Labs to harness and gain insight from previously unanalysed information. Our targeted approach to use of machine learning also created a much deeper understanding of the data, guiding the client on how to improve and personalise gameplay.

Both analytics dashboards - for the BfB team and for the healthcare professional - enabled improvement of the game to suit players needs. As a result of the work undertaken by Dae.mn there have been a number of enhancements in key areas:

Results

“We expect the ML module and associated dashboard to give us more information on player performance. This will help us measure the impact of the product for the first time and guide future development. The healthcare professionals who use the dashboard will for the first time be able to obtain analytics to monitor improvement. This will allow them to better support players and judge whether the game is having the effect it should be, whilst dramatically improving user experience and progress.”

“The work Dae.mn did for us will help us measure the impact of the product for the first time and guide future development - we will be able to dramatically improve the overall user experience.“

Director of BfB Labs