View the original post
We won ₹50000 at NES Innovation Awards with the Fitspace project
Fitspace was also selected for the grand finale of Toycathon India.
About Fitspace 💪🏻
Fitspace is an AR fitness application where you practice all your home workouts, recommended exercises as per user fitness goals along with a personal trainer all 3 in a single application.
Market Gap 🧩
Currently, many videos show how to do the exercise, but there are no apps that provide the AR environment regarding the same. It helps you to view the entire exercise in the real environment. The trainer in the app will be of the same height. The exercises will be curated, scheduled depending upon the various factors like equipment, number of days of workout, etc.
Target Audience 👥
Individuals who are unable to have access to gyms due to time, accessibility or whopping gym subscriptions are our potential users. People who cannot afford personal trainers or have basic fitness as their end goal are also our targeted users.
Product Potential 🥅
From the gym owner’s perspective, it is quite costly to afford gyms, trainers, equipment, electricity bills, etc. w.r.t. to the low purchasing power of the members.
🚀 Perfect time to go digital
So currently due to pandemics and work from home culture has made life quite lethargic and dull. 3 out of 5 people are not able to take care of their health and hygiene.
Every individual will think twice before going to the gym for working out with the fear of not getting any coronavirus symptoms. Those who are working out are still unaware of their workout schedule as well as the correct form of exercise and where one should feel exact pain while exercising. This app will help both of these categories of people.
Everyone does not intend to spend money on health and fitness as this is not considered a necessity, it is considered as a liability. But one can pay the reduced subscription depending upon his physique.
Going digital will help us cut down these prices to affordable ones. Where in a city like Mumbai one used to pay nothing less than INR. 10,000/- annually. By going digital we can provide a pizza pricing subscription of 399/- month.
We used various only sites that provide mocap data (pre-recorded stick figures). We built a 3D model in Maya Software and baked the mocap data with the 3D model.
The output of these 3D figures is scripted in Unity for the following height options, for design purposes as well. The final output will be in the form of an app.
User Survey 👀
We have collected surveys from various sources:
- Airtable Form
- Twitter Polls,
- Instagram Polls as well, keeping the Airtable form as our main source of data gathering process.
- Product Info
During the survey and 1–1 conversations with more than 15 individuals, we came across some insights which helped us understand the product better, to which we are currently headed as of (02/06/2021).
Unusual Tech Stack 🎁
The problem faced by us: For recording the exercises, one needs a tracking suit which costs around $2500/- It is not a viable option for a BE project.
- App development platform: iOS
- For making 3D models and baking the Mokap (Stick figure): Maya, Precision Neuron Suit, Axis Neuron Application
- To provide a marker-less AR environment to the animated 3D trainer: Unity 3D
- To design the complete app layout of our IOS applications: Figma
- Recommender system: Sklearn / Tensorflow
- Database integration: Firebase
Key Functionalities 🎯
1) It will showcase the key pain points on the body, on which part one should feel pain while exercising.
2) Curated exercise depending on various factors.
3) Diet plan, one-one consulting.
4) One can rent pieces of equipment, supplements from the same app.
5) Real sized AR trainer showcasing exercise in a real environment.
Product Journey (till date 20/09/2021) 🛣️
2) We have partnered with fitness professionals, recorded animations of more than 150+ exercises, prepared our data set as it has many constraints that need to be taken care of.
3) We have partnered with a smart glass company and publishing our product for their platform.
4) Further we will be working towards 3D Models animations and data modelling process, and then towards application interface, and building.
Future Prospects 🔮
With machine learning algorithms (object recognition) and few sensors, we can develop a product (a scanner) that can detect the motions of the body while exercising and guide him/her with the help of text output.
There are few more constraints regarding lighting, cost of implementation, full-body detection, etc. which is not put to practice to date.
We have also prepared a prototype, a pitch deck that includes the plans, if you like to have a glace at the same, do ping me.
Subscribe to my newsletter to know more about such interesting projects.
LinkedIn 💌| Thanks in advance 🙂