With the rise of affordable wearable technology, Australians have more opportunities than ever to monitor their health and activity levels. At the same time, advancements in machine learning (ML) and behavioural sciences have unlocked new ways to provide personalized, data-driven nudges that can engage users with the benefits of an active lifestyle. By combining these technologies, there is immense potential to support at-risk* Australians in forming sustainable exercise habits and reaping the long-term benefits of regular physical activity.

Our client is a team of researchers at a leading Australian university, who sought to create and develop a physical activity assistant that can be fit for the average consumer while generating data for future research. Their goal was to create an intelligent system that not only tracked physical activity but also engaged users with meaningful conversations, timely nudges, and ML-driven insights.

This case study explores how we collaborated with the research team to bring their vision to life through the following key objectives and deliverables:

• A mobile app where users can customise their experience and engage with the conversation agent. The app will also be a place where users can set goals, feel rewarded, and be given educational resources around physical activity.
• A calibration period where we learn about the user and their current relationship with physical activity.
• A nudge engine that notifies users to be active. Depending on the user's response or effort, the nudge engine will be updated according to the user's engagement.
• A conversation agent developed through conversation design, personas and focus-group testing. This agent is designed to adapt to the user's variables such as their lifestyle and environmental conditions.
• The design of a chat assistant character that can be customised according to the user's level of motivation or learning styles.