Skip to content

Meet Insane AI, a health app that tracks your physique movement utilizing cellphone’s digicam

Two years in the past, when Anurag Mundhada and Jayesh Hannurkar, graduates from IIT-Bombay, have been combating weight points in the course of the first wave of the coronavirus pandemic, they considered bringing their experience in sensors and laptop imaginative and prescient to create an interactive house exercise platform.

“The largest downside assertion from a shopper standpoint that we’re engaged on to resolve is motivation,” Mundhada defined the concept behind their Insane AI app. “Tremendous match fashions and trainers can generally actually flip off novices however should you make the preliminary onboarding easy and frictionless as potential, health will be enjoyable and is for everybody,” he stated.


Insane AI, which is presently in beta for effectively over a yr and nears official launch quickly, is totally different from different health apps, declare the founders. Not like Apple Health Plus or Nike Coaching Membership, Insane AI makes use of your smartphone’s entrance digicam to trace your movement throughout a exercise, they stated.

“The digicam tracks the physique and divides it into 21 factors: your shoulders, elbows, ankles, and knees. As soon as the monitoring of factors is completed in real-time, we run a video AI layer on prime of that which analyzes the movement and clarifies which train is being carried out, counts the reps, after which the system offers a rating of how intense the train is,” Mundhada advised indianexpress.com throughout a video name.

Anurag Mundhada (left) and Jayesh Hannurkar (proper) co-founded Insane AI in 2020.

“If you’re doing a squat and your hips are under your knees, you get a rating of 1 and in case you are simply doing very shallow reps, you’ll not get the identical quantity of rating,” he added. Mundhada claimed its AI-powered exercise monitoring expertise is 98 per cent correct primarily based on the corporate’s inside check knowledge.

The intention to make use of a smartphone as an integral a part of your health journey and create a brand new consumer expertise round cellphone cameras comes from Mundhada and Jayesh’s earlier expertise in digicam sensing applied sciences. Jayesh was with Sony Japan earlier than co-founding Insane AI and labored on LiDAR sensors, a expertise that provides depth scanning for higher images and AR in smartphones and is predominantly present in Apple’s high-end iPhones and iPad Execs since 2020. Mundhada, then again, labored with a Bengaluru-based Axogyan AI the place he did a challenge with Mercedes Benz for human sensing for autonomous autos.

Mundhada stated he and his group have been fascinated by Nintendo Wii and Microsoft Kinect and wished to convey the power to trace physique movement utilizing a cellphone digicam. Though smartphones through the years have gotten higher {hardware}, depth sensors are nonetheless restricted to a handful of pricy telephones. “Utilizing AI we enabled the conventional RGB digicam to trace real-time exercise monitoring which is current on all telephones to do what Nintendo Wii or Xbox Kinect can do,” he stated when requested how a smartphone digicam can observe physique movement with no further sensors .

The Insane AI app makes use of your smartphone’s front-facing digicam to trace your physique movement.

To check whether or not the smartphone digicam actually tracks physique movement and whether or not an app will be created round house exercises, Mundhada developed Pushup dojo which went reside within the early days of the pandemic. The app was used as a proof of idea to reveal that expertise and exercise monitoring utilizing a cellphone digicam could possibly be potential. It was a primary exercise app that garnered 2000 customers, largely buddies and instant household of the co-founders. That gave Mundhada some confidence to remodel the algorithms and consumer interface and ideal the physique movement monitoring expertise to create the Insane AI app in early 2021. This time, the group labored with health trainers to create the app and supposed a large consumer base.

Mundhada calls Insane AI a house exercise app with a “extremely gamified expertise.” The app, which will be downloaded from the Google Play Retailer and Apple App Retailer, presently requires individuals to sign-up for the waitlist, and starting the final week of October, Insane AI will grow to be out there to those that are presently on the waitlist.

To make use of the app, one has to first arrange a cartoon(y) avatar which will probably be your digital id on the exercise platform. You’ll be able to simply place the cellphone in entrance of you at a distance, begin exercising, and the digicam then begins monitoring your physique movement in the course of the exercise. Each exercise or train customers do will probably be counted and they’ll get rewarded by means of cash which will be redeemed within the type of digital collectibles and coupons for which the startup has tied up with a number of companions. Customers can even problem their buddies each in real-time in addition to turn-based health battles. The Insane AI app works on any smartphone that has been launched in 2016 and after. Within the coming days, customers will see a brand-new interface and a number of options together with much more exercises and workouts.

Mundhada calls Insane AI a house exercise app with a “extremely gamified expertise.”

The AI-powered exercise expertise mannequin developed by Mundhada and his group not simply tracks physique factors but in addition tracks customers’ health ranges throughout 5 totally different dimensions similar to power, cardio, higher physique, and decrease physique. Mundhada stated they’ve taken quite a lot of enter from health consultants to develop the exercise plans and resolve on totally different health ranges, whereas the corporate’s in-house knowledge science group has labored on making a health mannequin which will be personalized and customized as per a consumer . The insane AI app presents 35 totally different workouts and 150 exercises, out of which 4 exercises are free, to start with.

Based on Mundhada, Insane AI appears to be extensively used at 7:30 am and one other peak comes within the night, round 6 pm. On common, customers spend 9 minutes on the app, up from 5 minutes within the early days of Insane AI. Nevertheless, energy customers spend nearly an hour on the platform.

“Insane AI will not be meant to switch a private coach utterly,” Mundhada stated, including the app is “designed to have a straightforward pathway to studying health and getting extra superior.” “For circumstances similar to excessive weight problems, it’s all the time advisable to have a human within the loop and that is not only for studying but in addition for the prevention of accidents somebody who might help you notice you whenever you’re lifting weights.” Mundhada is obvious that Insane AI goals to supply a rewarding and enjoyable expertise and complement your journey alongside the best way.

The house exercise app comes with a promise of not amassing customers’ footage and movies however solely amassing anonymised knowledge of movement itself to enhance the AI ​​engine which does the movement monitoring.

For monetisation, Mundhada says they’re going for a freemium mannequin that can let customers expertise the app totally free, with sure options similar to superior challenges and redeeming rewards requiring a subscription. Proper now, those that sign-up for the app will stand up to 6 months of free subscription. Mundhada and the group are but to determine what could be the price of a subscription.

Final yr, the start-up raised $900,000 in seed funding. The spherical was led by early-stage enterprise capital fund pi Ventures, Anupam Mittal, Sameer Pitalwalla, and different angels in sports activities, health, and tech. In the long term, Mundhada needs to show Insane AI right into a health metaverse and work on constructing a basis for an immersive world that has already been began.

.

Leave a Reply

Your email address will not be published. Required fields are marked *