Physimax CEO Ram Shalev: Facilitating Pro-Level Training with Care for Every Active Individual

How Physimax has built real-time scoring and optimizing musculoskeletal conditions with Computer Vision and AI.

Yuliya Sychikova
COO @ DataRoot Labs
23 Nov 2020
6 min read
Physimax CEO Ram Shalev: Facilitating Pro-Level Training with Care for Every Active Individual

Ram Shalev is the co-founder and CEO of Physimax, an AI-powered musculoskeletal self-monitoring and optimization.

Founded in 2012, the company raised Seed funding from Lionbird, NetCapital, Beyond The Game Network, SportDocs, and SportGurus. In addition, the startup is a graduate of the Microsoft Accelerator, Tel-Aviv and Georgia Tech Flashpoint Accelerator program, Atlanta.

Based in Tel Aviv, Israel, Physimax developed and validated a computer vision and artificial intelligence software technology to capture, model, and evaluate body movement, initially addressing athletes for injury prevention, performance development and return to play. In just a few minutes, using any Android / iOS smartphone, and with no-wearables, Physimax cloud service objectively scores musculoskeletal condition (mobility, joints’ stability, strength), monitoring & optimizing musculoskeletal conditions. Depending on the musculoskeletal profile, the application recommends suitable athletic training program.

How was Physimax born?

As a high-performing individual with a MSc degree in engineering and a wealth of experience in developing innovative software services, together with David Kahani my partner – an avid sport analytics fan, we identified and validated a market lacuna - scoring functional movement objectively and in real-time, using software-only-AI with no physical markers nor wearables. Our initial addressable market were athletes and we aimed to educate them about a) risk of injury, b) performance development, and c) recovery progress and readiness for return-to-play. Physimax validated technology reliably scores mild movement limitations that derive from musculoskeletal deficiencies which can be altered by workout prescirbed by service. Our vision is to facilitate pro-level training with care for athletes.

Physimax is providing an individualized training program based on a sophisticated real-time personal analysis of human motion. How does your Computer Vision and Machine Learning technology work?

Physimax technology combines multiple layers.

  • First, identifying & tracking body motion in 3D or any 2D video (with no markers / wearables), validated for accuracy and robustness.
  • Followed by extracting the dynamic movement characteristics, scoring movement deficiencies ("errors") as-if a group of experts (for objectivity reasons) watched the individual and scored aspects of: Mobility, Strength, Stability, Technique and Performance.
  • Providing in-depth evaluation of the body functional capacity e.g. knee stability, feet strength; scoring feet to shoulders, comparing the functional performance to the relevant group according to: activity type and level, gender and age group. Physimax developed a robust engine that can be trained to fulfill any evaluation protocol a human expert can visually score but better with immediate access (i.e. by smartphone) and feedback, which is objective, repetitive and provides actionable recommendations. In addition, it highlights the most probable cause for the deficiencies score e.g. weak muscle group, limited stability etc.
  • These Insights are followed by weekly workout programs that contain specific exercises that address the findings. The service selects the training scope from activity libraries defined by customer, so that the outcome is sport specific workout programs.

ML/CV models require lots of data to train the algorithms. Where did your data come from and how did you approach the training?

Indeed, the initial set of normative values per a specific group of athletes is based on hundreds of scientific publications we manually reviewed and provisioned to the service. The service collects the motion data samples performed by the end-users, the normative values are refined according to the data samples collected from a relevant group of monitored individuals. To-date normative values were refined in a variety of sport types: soccer, basketball, football, baseball, with different age groups: youth, high-schools, collegiate and professional level, during seasonal training and after-injury rehabilitation.

The challenge was to build a tracking algorithms that dissect complex movements into measurable segments for the Machine Learning algorithms to compare with normative values and score the movement objectively as if a group of experts are watching the video.

Ram Shalev

CEO @ Physimax

What were the significant-tech challenges that your engineers had to overcome while building Physimax?

The major technology barriers were obtaining high measurement accuracy while utilizing "commodity sensors" using computer vision analytics, setting reliability algorithms to reliably track the dynamic movement. With home capturing unlike lab capturing the accuracy measurement varies and the algorithms should weigh different the frames, setting high scores to the most accurate frames. The challenge was to build a tracking algorithms that dissect complex movements into measurable segments for the Machine Learning algorithms to compare with normative values and score the movement objectively as if a group of experts are watching the video.

To support the measurement and scoring accuracy (anyone can theoretically claim that they track and score movement), we have been collaborating with Tier-1 US & Israeli academies: University of North Carolina, University of Connecticut and the US Military Academy at West Point and Walter Reed National Military Medical Center to independently validate the measurements and scoring reliability, implementing evidence-based test protocols.

Furthermore, to make the service actionable and self-use, we developed learning algorithms that highlight the most probable cause of the movement deficiency, set a quantified digital musculoskeletal profile of the individual and recommends complementary exercises, in the specific context of sport type and level.

We developed learning algorithms that highlight the most probable cause of the movement deficiency, set a quantified digital musculoskeletal profile of the individual, and recommends complementary exercises, in the specific context of sport type and level.

Ram Shalev

CEO @ Physimax

Physimax is chosen by elite organizations, including NBA teams, D1 colleges, US Military units, top healthcare providers, and academies. Was there a particular client story that you can share where Physimax technology made a huge impact?

Being chosen to commercially serve elite sport teams like the Chiefs, Utah Jazz, Flamengo and Paris Saint Germain soccer teams as well as the Israeli and US army, established great credibility for Physimax. Without disclosing names, I can share that teams and athletes who have been continuously using Physimax decreased their non-contact injury rates by 60%, the reported ease-of-use and engagement have been very good as well. Note this impact is influenced by multiple parameters intrinsic and external with Physimax solution is a must-have to objectively and proactively monitor the athlete musculoskeletal functional performance, and hence - condition, scientifically impacting the winning ratio.

Teams and athletes who have been continuously using Physimax decreased their non-contact injury rates by 60%.

Ram Shalev

CEO @ Physimax

Since July 2020 Physimax has been used by Japan's Activate Gym for the rehabilitation of pro-athletes and training of young athletes. Is your solution tailored for rehabilitation or performance optimization or both?

As you share, the customer uses this for both, refining the terms - "the performance optimization" is actually athletic performance development that can serve the average person to functional improve, not less important is injury prevention or mitigation to identify deficits that under load are likely to end with an injury.

How did COVID-19 shake up the reality of your business?

The COVID-19 influenced our customers and potential customers significantly, as season scheduling was changed, tickets income impacted and “on-site” activities had been limited. However these challenges actually presented also a huge business opportunity - It forced us to push forward on the offering and engagement with providers serving recreational athletes, providing “at home” no touch solution available in a B2B SaaS licensing model that can be activated remotely.

What are your plans going forward in terms of product, team, and sales strategy?

From product and technology, we obtained a patent in the US to protect the unique IP and are in process of obtaining one in Europe. From business perspective, we are thinking of providing our technology to serve the mass market of recreational athletes and following the average person through strategic partnerships with digital platforms that reach out to millions of athletes and health-related services.

We aim to integrate our technology to serve consumers through strategic partnerships in the following fast growing markets: Home fitness, Home Physical Therapy and Wellness.

I believe the ultimate goal is to keep improving the AI to Deep Learning, naturally interpreting the end-user movement behavior, comparing, and leveraging the collective data analysis power.

Ram Shalev

CEO @ Physimax

What do you think is next for AI/ML + IoT in Sports and Rehabilitation?

Wow, a tough question, I’ll try to simplify the answer. I believe the ultimate goal is to keep improving the AI to Deep Learning, naturally interpreting the end-user movement behavior, comparing and leveraging the collective data analysis power. Physimax vision is to democratize direct access for every active individual (3rd generation as well), to achieve the personal functional performance goals.

Have an idea? Let's discuss!

Book a meeting
Yuliya Sychikova
Yuliya Sychikova
COO @ DataRoot Labs
Do you have questions related to your AI-Powered project?

Talk to Yuliya. She will make sure that all is covered. Don't waste time on googling - get all answers from relevant expert in under one hour.
OR
Send us a note
Optional
File requirements pdf, docx, pptx

Author

Yuliya Sychikova
COO @ DataRoot Labs
Yuliya is a co-founder and COO of DataRoot Labs, where she oversees operations, sales, communication, and Startup Venture Services. She brings onboard business and venture capital experience that she gained at a leading tech investment company in CEE, where she oversaw numerous deals and managed a portfolio across various tech niches including AI and IT service companies.
offices map
Kyiv (HQ)
Max Frolov
CEO @ DataRoot Labs
Tel Aviv
Ivan Didur
CTO @ DataRoot Labs
Los Angeles
Yuliya Sychikova
COO @ DataRoot Labs
builds and implement AI-powered systems across different verticals to help our clients operate effectively.
GoodFirms Badge