Workout IoT app.

Key Idea of the Product

A mobile app collecting data from a bunch of sports equipment devices, and transforming this data into valuable personalized recommendations and a virtual portable private gym.

Goals set to RedCat

• Collect and prioritize the requirements of the MVP version of the product, and develop internal calculation algorithms.
• Sagasiously develop app DB architecture in a way it’s customizable and compatible with AI and BigData in the future.
• Connect with a bunch of physical devices which are not available to transport to us.

Constraints and Challenges

Pressure from the Client’s investors, missing physical devices, extremely high requirements for the device connection stability, complicated app back-end architecture.

We started with a very intensive 3 weeks Discovery phase to cover all possible logical gaps and collect candidate release requirements for V. 1.0 and V 2.0. Once the Client adjusted the design files accordingly, we moved further with 2-week Scrum iterations.

As a result RedCat team:

• established bi-weekly app release process for Delivery manager and marketing team;
• managed to connect to the hardware device remotely and still test it;
• created a microservice architecture and back-end logic in a scalable and autonomous way for adding ML algorithms;
• the front-end development approach was created in a way it requires minimum effort in case of future design changes.

Delivery strategy

The work was performed in the Client’s repositories and infrastructure, which ensured daily updates. We had demos each week, and also some Scrum rituals to work through change requests and Client feedback.

Unique features

• Ability to connect to a bunch of custom devices – special sports equipment with pressure, acceleration, and tension sensors.
• Each repetition in each exercise is measured by sensors, which means users get fair feedback on their performance during a training session.

Mobile app features

• Enormous collection of exercises with all kinds of sports equipment offered by the Client and foal-based suggestions in training plans.
• Personal records and statistics are collected directly from IoT devices and transformed into a visual representation and ranking among other app users.
• Individual custom training plans and challenges.

Data management

• The data from sensors is collected 10 times per second.
• At the end of each repetition, and a whole exercise app calculates values to be displayed to the user.
• The raw data received from sencors is compressed and stored on the backend for a future training of an ML module.