Treating Workers Like Athletes to Prevent Injuries

Pete Miller [00:32] Welcome to Predict & Prevent! I’m your host, Pete Miller, CEO of The Institutes.

Today, we’re joined by Sam Hollenbach, CEO and President of Amp Health — the movement health company on a mission to help people move better, prevent musculoskeletal injuries before they happen, and speed up recovery when they do.

Sam brings a unique perspective to workplace safety. A former college and NFL quarterback, he went on to earn an MBA from Wharton, spend over a decade at Under Armour leading global digital product strategy, and eventually channel all that experience into reimagining how we protect the modern workforce.

Amp Health’s vision is to provide its customers’ employees with the same kind of personalized physical preparation as an elite athlete, minimizing strains and sprains that can lead to injury.

In this episode, Sam walks us through how Amp Health uses computer vision and movement science to deliver a fully customized, six-minute pre-shift warm-up for every employee. This program is built around each worker’s individual mobility, job demands, and injury history.

With musculoskeletal injuries a leading health care cost for employers, what Sam and his team have built just might be the game-changer the industry has been waiting for.

Pete Miller [01:57]: Well Sam, welcome. to get things started, can you share with our audience a little bit about what AMP Health does and kind of, you know, give us the elevator pitch if you will, and then we can get into more detail about how you do that.

Sam Hollenbach [02:11]: Yeah, sounds good. Thanks, Pete, for having me. Amp is the movement health company. We want people to move better, and we believe that better movement helps people avoid pain and injuries, and also helps people recover faster from pain and injuries when these occur. We have two products in the market, Amp Prevent and Amp Recover. And the Prevent product I’ll be spending more time about talking about today. but that started as a purpose-built EMR platform designed for on-site occupational health professionals. And then we’ve sort of made a bunch of additions to it. We’ve added to it with some pretty innovative features, which I’m excited to talk about that really specifically focus on injury prevention.

So, one of these sort of recent products that we’re excited about is called workforce readiness. We think that workforce readiness really represents a step change in injury prevention and worker wellness. Workforce readiness starts with a belief that a pre-shift warmup is an effective tool in reducing injury-related KPIs like lost days, et cetera. And there is some literature that has demonstrated that the causal effect there. However, what we have seen is that it can be a one-size-fits-none. That’s what we call it internally, type of scenario. So that pre-shift warm-up is hard to get specific for every employee that’s showing up at the job that day. There’s people move differently, people have different job demands, people have different injury history, all of those things are related to your likelihood or your risk of a future MSK injury.

And so we developed the world’s first fully customized pre-shift warm-up. And that’s what we’re calling workforce readiness. It starts with a benchmark of each employee’s mobility through a three-minute computer vision assessment. And then also we ingest job demand and past injury history for each employee, which gives us a really good sense of where that employee’s individual risk is. then we built a model which we developed in partnership with University of North Carolina, their School of Exercise and Sports Science. And this model assigns specific movements, basically builds out a warm-up stretches and exercises for each employee. It’s designed to target each employee’s individual risk profile. And the output of this is a six-minute warm-up that each employee would follow on their phone.

Everybody’s getting a different plan according to their own needs. And so every month each employee is supposed to take a new computer vision assessment, sort of get a fresh sense of where their risk is, and then is delivered a new set of warmups. So that’s workforce readiness. We’re really excited about that product. We think it’s a novel approach to injury prevention.

Pete Miller [05:36]: Yeah, sure is. So before you joined Amp Health, you played college football, you played quarterback at the University of Maryland, and then some in the NFL for Washington, right? So before then transitioning into engineering and business. So how did your experience with elite physical preparation in sports inspire the mission of Amp Health and your concept of the industrial athlete?

Sam Hollenbach [06:00]: There’s a lot of parallels there. I think probably the most fundamental of which is that in both sport and in industry, your body is your tool. It is the tool. Your body is critical in doing your job as an athlete. And you know, whether that’s in sport, whether that’s in industry. So that general idea of preparing your body for work, caring for the body after work. Yeah, those concepts I think carry through pretty well across both. That did help inspire our thinking around workforce readiness. As a football player, you’re not going into a practice or a game without a warm-up. And likewise, you are, you’re not going to necessarily be doing the same workout. A kicker and an offensive lineman, they’re not doing the same things to get ready for their job.

So that was part of the you know part of the fundamental understanding that we that helped us think about workforce readiness. We did a lot of digging, a lot of research with some of the top movement science experts. I mentioned UNC earlier. And we found that this idea does in fact translate well into industrial jobs. Of course, the scale is very different. You’re going from, you know, a football game being a very intense, sort of compressed period of activity once a week for four months a year or whatever, to a job that you’re doing sort of chronically every day, lower intensity but higher volume over time. but the Idea does translate really well. Warm warming up, being ready, relevant warm-ups, really helping people get ready for their job. And so that’s I think one of the main ways that sport and industry are connected. from a you mentioned mission, from a mission perspective, wanting all people to move better. Better movement is very central to the journey of every athlete. Certainly was part of my experience. You know when I think about better movement I’m not necessarily thinking about injury prevention. I’m really referring to and thinking about less pain, less discomfort associated with my job so that I could stay available. Every football player they just move a lot better in August than they do in December. Ha ha, the season just beats you up.

And again, it’s a different scale, but years and years of physical work, you know, showing up doing a job on the factory floor, that can produce these kinds of similar movement limitations that could cause some pain, cause some discomfort. and so that that idea of helping to reduce discomfort through better movement is very central to you know, was a big part of my experience as an athlete and a big part of what we built into workforce readiness as well.

Pete Miller [09:05]: Yeah, I think I mentioned I played college football too. I get I get the preparation, couldn’t agree more, could not agree more. So we know that behind sort of cancer, musculoskeletal injuries are the number one health care cost for employers. So beyond preventing these injuries and their cost, how does Amp Health’s mission focus on a worker’s holistic well-being, and their quality of life actually, after they clock out. How does how does that all work together?

Sam Hollenbach [09:37]: There’s an implicit focus on the end employee because our pre-shift warmups are really designed to target areas of the body that are at the most risk. And the model that we’ve built, there’s a correlation between areas that are at the most risk and, you know, joints that are sort of feeling the most discomfort. and so we are sort of fundamentally giving the employee movements that are designed to improve their mobility and really target that discomfort. Even with that in mind, you know, as you’re using this product, as employees are using this product over time, they may wake up one morning and just feel a sense of discomfort. You know, everybody kind of knows what that feels like. You wake up, your back hurts, your neck hurts, something like that.

We have a feature in the app that allows an employee to indicate that to us. They, you know, I feel some pain in my neck. I feel some discomfort in my neck. and if we get that input from them, what we have built into our product is the ability to give them in that morning’s warm-up some specific movements for that area of their body that is feeling some discomfort. So there’s kind of the overall plan that employee would be following that’s associated with their mobility, their job demands, their past injury history. But there’s also a more sort of focused and targeted sort of near-term over a couple of days component to our product that allows employees to tell us when they’re feeling some discomfort and then have some movements that are designed to reduce that discomfort.

So that’s one of the ways that we focus on employer and sorry, employee well being. I will also add that we have a weekly survey that is built into our app. It’s asking about employees’ energy level, muscle soreness. There’s a few other things in this survey. And we believe that this data aggregated over time can help show employers where there could be some trends that potentially show maybe where there’s some systematic issues that could be solved systematically to help reduce some of those negative outcomes, like poor soreness or high soreness, poor energy associated with certain jobs or in certain departments or with certain seasonality.

So overall we have a few features built into our app that allow us to collect some data that we can then bring into you know an employee experience that helps them reduce their pain and discomfort, that gives their employer some insights into ways to maybe improve the working conditions potentially to allow for an improvement in workforce wellness. And that all really stems from a real focus on trying to build a real user-friendly app, employee friendly app, that is sticky, that keeps them using it and that, you know, keeps delivering this kind of insightful data.

Pete Miller [13:03]: You know, one of the things I found interesting, because when you think about it, traditional ergonomics sort of, you know, changes the job to fit the worker, for example, by adjusting shelf heights or limiting box weights. But it seems like you flipped that, right? So that your approach focuses on workers’ physical readiness for the job. So why is assessing an individual a more effective approach than just redesigning the task?

Sam Hollenbach [13:30]: That’s a great question. This is this comes up in many of our conversations. You know, we ultimately feel that measuring the individual is the best approach. It certainly isn’t you know, taking the place of looking at the job and the engineering approach to the job. But one of the reasons we really like it is because it does account for the other sixteen hours of the day.

You know, an example that I like to use is if someone shoveled their driveway, you know, you got snow last night. You’re out there shoveling your driveway. That is I do that at home. It’s always so hard. It’s always I come in, I’m beat up after that. And this is impact on the body that our computer vision would detect. We would, you know, you’d show up for work that morning in a different readiness state with some different risk factors than if you did not shovel your driveway. So we would account for that in the way that you’re moving. We would give you movements. We would give you a warm-up that is accounting for that, that is helping you recover from the impact of shoveling your driveway. That’s at a high level why we really like the focus on the individual. And then the other thing too is, you know, optimizing for a job can get tricky from an anatomical perspective. You know, you have different types of anatomy. I’m six foot five. You know, the risk factors associated with me doing a job are different than the risk factors with someone who’s shorter doing the same job, which of course there’s tools to measure me versus someone else doing that, but you know, it’s something that you would always have to be doing is sort of measuring every person going into that job. What we like to do is just focus on the individual, have a longitudinal view of the individual’s movement, which is going to reflect the demands of their job. So the way that they’re moving, the mobility that they have, the soreness that they have, I mentioned the surveys earlier, all of that stuff is going to sort of be representative of the job itself. So we feel that kind of overall, a focus on the individual is a more holistic, a broader approach that in the long term is going to show data that can be more, more actionable for that individual.

Pete Miller [16:03]: Yeah, that just makes a whole lot of sense. As somebody who likes to shovel his driveway Sam every the next day, I’m like, yeah, it was a good workout, but I can’t move. That’s problematic.

Sam Hollenbach [16:15]: You’ve got to be careful. Use the legs, lift with the legs.

Pete Miller [16:21]: Use the legs! Exactly. You mentioned earlier that Amp Health starts with what you call the three-minute movement health assessment. So just can you just walk us through that from the workers’ point of view? You know, what do they actually do on the floor and like what equipment is involved?

Sam Hollenbach [16:39]: Sure. Yeah. As an employee, you would be logging into the Amp Health app. There’s, you know, if you’re supposed to be doing a movement health assessment that day, it’s very clear in the app. You’re going to, you know, tap the button and you will be shown what to do. What the employee is essentially doing is setting the phone up in you know a way where you can look at it and you’re taking a few steps back. We’re going to use we’re using the front-facing camera of the phone and walking you through six different movements. Again, it takes three minutes. You have to position yourself in a few different ways before you do a movement. So what we convey in the app is this is like a clock. If you’re standing and you’re looking at the phone and the phone is 12 o’clock, we may ask you to position your feet at 11 o’clock or at 1 o’clock. And then we’re showing you exactly what movement to do.

The movements are representative of many types of industrial jobs, the demands that you would have in many types of industrial jobs. One of them, for example, is a is just a squat. So as you’re squatting down, you know, we’re picking up a lot of data that can help us understand your mobility among other things. There’s no additional equipment that’s needed. You’re walking through this and as soon as it’s done, everything is saved and you’re ready to go. And so from an you know it’s from a time perspective, it’s not a big ask. It you know takes three minutes and we just get a lot of rich data from it that’s informative to your warm up.

Pete Miller [18:20]: So how do you analyze that data? Like what technology is picking up, you know, that data and maybe things that even and, you know, what does it analyze, what does it see, maybe even things that an experienced safety manager watching the worker would miss?

Sam Hollenbach [18:36]: One thing is it’s highly accurate. So you know, that’s I think important to note. It’s accurate in different, you know, mobilities. So you know, certain movements we’re going to look with high accuracy at what your mobility is. It’s also precise test over test. I think that that’s another important piece of this, you know, precision gives you the ability to better understand workforce wide, where there could be musculoskeletal risk concentration because you have the same test, the same assessor essentially used for every employee. It also is valuable that precision is valuable over time. It’s valuable to show relative changes in an individual, you know, to show where. you know, risk could be risk could be decreasing or increasing.

But back to the accuracy you know part that I mentioned earlier, we’re also measuring asymmetries and compensations. That’s something that’s difficult for maybe a you know an experienced individual to look at from a naked eye. Compensations we would define as the you know the movements that you have to take as an employee or as an individual to put yourself in a certain range of motion. An example would be one of our movements is a shoulder rotation. And you know, we’re going to look at how far you can rotate your shoulder, but we’re also looking at do you have to move your neck to reach your maximum range of motion in your shoulder. And that is a that is a sign. That neck movement would be a signal of a compensation we would pick up on, we would recognize that as a as a movement deficiency in your shoulder that we would then target downstream. So the computer vision is really good at picking up all these things. And again, doing so in a way that scales well very quickly and rapidly across a big workforce.

Pete Miller [20:39]: So anytime you use a camera with employees, right, privacy always seems to come up, especially in workplaces that have a union. So how does your approach address those concerns? And like what data does the employer actually see?

Sam Hollenbach [20:58]: So first of all, the video, we’re not we’re not keeping any video. We’re doing all of the compute on edge, meaning on the phone. So all the algorithms are computed right there. And the results are computed on the phone. We’re never storing or sharing video. And that gets us I think 90% there from a privacy perspective. Cideo is a big deal. So we’re not we’re not sharing, we’re not storing video. You bring up a good point in unionized work workforces, you bring up a good point about sensitivity on what the employer is seeing. All of that is a sort of an optional configuration for our customer. so many of our customers do not want to see that data. They just don’t want to see it. And so, of course, there’s never any videos shared or stored, but the actual employee risk data. So, you know, what joint, what does an employee’s movement health assessment look like? In often in in many cases, we are not going to show that to the employer. They just won’t see it. We can do that because we have a fully end-to-end solution. So that employee’s data can go right into that employee’s warm-up. And we can just continue to leverage computer vision data and all of their risk factors in a way that is insulated from the employer visibility, but continues to give that employee the exact sort of optimized warmup for them.

Now, one thing that we do we do see employers like to get access to is the aggregate level of data. And so, you know, that would include things like, hey, at you know, plant A, shift B, show me, you know, show me, you know, what the highest level of risk looks like. or in this particular job family, inbound logistics, you know, what is my biggest level of risk there? And that is something that our product can show. It does so in a way that still really preserves privacy for the individual employee. No, you know, no names are shared there. but also gives the employer the ability to take action if they if they want on that data. So that’s a good that’s a good compromise. We see several of our customers opting for that route.

Pete Miller [23:21]: So suppose that the employer takes a video or the employee takes a video of themselves and they find a gap, like you mentioned shoulder rotation. So how does AMP Health sort of take that, you know, and then how does it how does it take that and then build the six minute sort of warm up? Like how does the engine combine the assessment results with the task that the worker’s going to do and their injury history to kind of design that routine?

Sam Hollenbach [23:46]: So, it’s sort of a sophisticated set of models that are doing that, but at a very high level, what we’re doing is segmenting out those six minutes. I mentioned it’s a six-minute warm-up. Each of those minutes represents a separate focus for our algorithm and does is designed to target, you know, the objective for that minute of programming, if you will. So, we’re going to use the first minute and we’re going to look at the biggest movement deficiency that we saw in computer vision. We’ll use the second minute to look at the second biggest movement deficiency that we saw in computer vision. We will use some of the other minutes for your job demands. job demands are entered by the employee. So and you know, we know that employers often have job demands. Employee, there’s value in the employees’ perception of the demands of their job. That is what we ingest on our app. And that gets programmed into a minute of programming. And likewise, previous injury history as well. If we see that you’ve had a big injury in your shoulder in the last, you know, couple of years, we’re going to use a minute of programming in the warm up to focus on that shoulder, to help rehab your shoulder, to help make sure it’s you know it’s being targeted.

So at a high level that is that is what’s happening. Now there’s a lot of sort of science in things like, you know, how deficient is the gap. You mentioned a mobility gap. How big is that gap? And that can dictate, for example, how much programming, how many minutes of programming in a warm up you would be given to address that mobility gap. But at a high level, yeah, we’re just we’re really sort of segmenting out this warm-up. And as an employee, you’re not feeling any of this. You know, you’re just kind of following along. I didn’t really talk about the employee experience of the warm-up, but it’s a real sort of set it and forget it type of experience. You’re pressing play and you’re mindlessly just kind of walking through what you’re seeing the model on screen on our screen on our app telling you to do, showing you what to do. And so you’re able to kind of just go through that pretty, pretty mindlessly. but there’s a lot of science that’s delivering that content that’s used to drive that content and deliver that content.

Pete Miller [26:21]: So Amp Health recently became a finalist in the National Safety Council’s MSD Prevention Innovation Challenge, and you received a grant to pilot the technology with Compass Group Canada. So congratulations on that. Can you tell us a little bit about the specific hurdles you’re testing in that pilot and what data you’re hoping to validate?

Sam Hollenbach [26:44]: Yeah, thank you. We’re proud of that. We think that you know that recognition it is representative of you know this being an innovative feature. So we really we really appreciated that and are excited to work with Compass. You know, in terms of hurdles, the main one that comes to mind is the workflow, right? We recognize every employer has a work workflow. And this is doing something potentially a little different. It’s you know we’re bringing in a you know a new experience for people and asking them to do something a little different than what they would normally be doing. So, that is that is a hurdle and we’ve been working with Compass. They’ve been great to work with. They’re, you know, helping us understand what a typical day in the life looks like for some different plants, some different divisions. And we are doing some things to help reduce friction around implementation of our product into those different kinds of operating models. In terms of data that we’re hoping to validate, certainly we want to see a reduction in injuries related to our product. Now, that is less simple than it sounds. Everyone’s definition of an injury seems to be different, and it’s actually hard to get actual injury data. and so we’re going to see how that goes and what kind of data we can get around the injuries themselves.

But what we’re also really focused on and compass is aligned on this is some of those other metrics on workforce wellness. And I mentioned the survey that we have. That survey is really powerful. That survey gives us the ability over several months of running our product to show trends, to hopefully show increases in employee energy. To show hopefully decreases in muscular soreness sort of at the aggregate level. And so that’s a big area for us that you know we’re excited to hope validate as a result of this pilot.

And there’s one other thing I’ll mention related to this Compass opportunity. We’re getting the chance with this pilot to show an even newer product that that we’re excited about that is really a group application of workforce readiness. And so without getting into a ton of those details, essentially we’re just looking at every employee individually in the same way that we do for the individual workforce readiness. But then we’re grouping, we’re categorizing employees into similar risk groups and then providing a warm-up for them so they can just show up, don’t need their phones, you know, basically watch a TV where, you know, that TV is going to show bring them through a warm up that’s relevant to their, you know, their sort of common risk areas. So there’s a lot going on with Compass. We’re really excited about the opportunity and you know, excited to have the partnership with them.

Pete Miller [29:59]: Sam, I understand you have a product that’s targeted toward work desk employees. Can you tell us about that?

Sam Hollenbach [30:06]: Yeah, absolutely. We have a product that we call workstation optimization. And this this is something that we were excited about. We recognize that we’re spending a lot of time focusing on, you know, industrial jobs where there’s a lot of different physical demands. A lot of times these employees are also at the desk a lot. And then beyond that, there’s a whole division of the workforce that is desk employees. I am one, I recognize how I feel after eight hours of just sitting at my desk working. It’s not the best feeling. So we wanted to use our computer vision to help identify workstations that are not optimized, that are not optimal for that employee. Again, I mentioned I’m six foot five. So, you know, what you do in this product as a user is you enter your height and then you do a quick computer vision scan. This is on the order of seconds and not minutes. but we’re taking a full anatomical scan of your body. We’re looking at the joint distances, you know, things like shoulder to elbow joint, for example. And then we’re using a model that’s actually an open source model from the Mayo Clinic that defines what an optimized workstation should be given certain joint distances. And we’re delivering that instantaneously to the employees.

So you would then know how, you know, to address your workstation, maybe to raise or lower your seat height or your desk. But we wanted to take it even a step further than just showing what an optimal desk setup looked like. We also wanted to give employees who don’t who can’t reach an optimal desk setup. You know, there’s only so much you can do. I’m not sitting at an adjustable desk at this point. So, you know, I’m sort of sitting with that risk factor throughout the day. So we’ve developed concept called “movement snacks,” which really is just going to give me throughout the day a couple of moments where I’ll get a notification on my phone.

We like the idea of this as an Apple Watch app or Android watch app as well, where you can just sort of get a little haptic notification and hey, it’s time for a movement snack, and then give you a certain movement, you know, 30 seconds or a minute of a stretch that you can do that specifically targets you know, the risk that you’re kind of generating, the discomfort that you might be generating by sitting in a desk setup that’s it’s not optimized for you. We’re excited about that product workstation optimization. We’re seeing some real interest from that among our among our customers and among in within the market and excited to continue to invest in supporting the movement health of desk employees as well.

Pete Miller [33:15]: Yeah, that’s great. As a desk employee, I’d appreciate that for sure. I get that. So great.