Karl Zelik

💡 I wrote this 5 years ago about #prosthetics tech transfer. Rereading it now, I think the three lessons hold up extremely well:

Number 3 came up in a lab discussion recently & numbers 1-2 are more important than ever.

While I originally wrote this about prosthetics, these lessons apply more broadly to other wearable devices too, including #exoskeletons & #wearablerobotics.

🎯 3 Things I Wish I Knew About Tech Transfer Years Ago When I Started Doing Prosthetics Research

If I could go back to when I started grad school, here are three things I wish I knew about translating prosthetic innovations out of the lab & into society.

These insights are partially from my experience as a prosthetics researcher, device developer, & wearable tech entrepreneur. And they are partially from a roundtable discussion I co-led with MIT Professor Hugh Herr at the American Society of #Biomechanics conference in 2021.

1️⃣ Reimbursement is often the major obstacle, not FDA approval

FDA approval or clearance is critical for bringing many new medical devices to market. But that does not mean insurers will cover or pay for them.

Reimbursement is a whole separate mountain to climb, potentially involving coding, coverage & payment.

This makes it incredibly hard for medical device startups when there isn't already an existing reimbursement pathway.

2️⃣ Bionic (aka wearable robotic) tech often takes over ten years & tens of millions of dollars to commercialize

This timeframe & cost to commercialize are based on Prof. Herr's experience commercializing several bionic technologies [and also align with my own observations, experiences & conversations across the wearable and assistive tech fields].

Herr warned that successful tech transfer takes grit & perseverance. He tried to start his bionic ankle company twice, failed both times, & only succeeded on his third attempt.

3️⃣ Creating low-cost prototypes is NOT the same as creating low-cost solutions/products. Make sure you know the difference.

This is a common misconception in academia. I frequently see university stories (often involving 3D printing) touting low-cost prosthetic devices. But these are often more fiction, or wishful thinking, than factual.

If you're looking to create & commercialize low-cost solutions, then you need to understand that the cost to build a prototype has very little to do with the cost to manufacture, maintain, or service a product at scale.

☝🏽 That point is worth reading twice ☝🏽

The latter cost is what you need to understand. The former is mostly a distraction.

Keep innovating. Keep pushing the bounds of science. There are so many people with limb loss & other physical disabilities who could benefit from advances in prosthetics, exos & other wearable tech.

But go into R&D with your eyes open. The more we understand tech transfer in academia, the more we can do groundbreaking research that becomes accessible & impactful in society. 🚀

🤦🏽‍♂️ Social media has given people a pretty distorted idea of what recreational hip #exoskeletons actually do. Traditional media often hasn’t done much better.

I’ve been using a hip exo for a few months. The simplest way to explain what it does:

It’s basically a leg-swing machine.

The exo's motor assists hip flexion during swing phase: helping bring your leg forward & up each time your foot is in the air.

It provides little or no hip extension assistance during stance, when your foot is on the ground.

I find the leg-swing assistance most helpful when walking fast, uphill, or upstairs.

Much less useful when walking slowly, downhill, or downstairs. I’ve also been surprised by how little it seems to help during jogging.

💡 One of the most interesting things I realized:

By focusing primarily on leg swing, the control problem is massively simplified.

I believe this is a major (and under-discussed) reason powered hip exos are proliferating faster than powered knee or ankle devices for recreational use: it’s simply an easier wearable robotics control problem.

☝🏽 A few other surprises:

1️⃣ It’s very quiet. If an exo makes enough noise to draw attention everywhere you go, people won’t want to wear it.

Quietness matters more for adoption than another few watts of power or Nm of torque. Exo developers still tend to think more power is better. It definitely is not with these hip exos.

2️⃣ It’s lousy for bending/squatting. Not only does it provide no assistance for these tasks, its frame physically gets in the way and causes discomfort or movement interference for me. This also means I can’t sit comfortably in most places while wearing it.

I knew this wasn’t the intended use case, beyond helping with the walking (while carrying) portion. But there are photos/videos online of these recreational hip exos during lifting & carrying tasks. Which is confusing. This device was even less compatible than I anticipated.

3️⃣ People don’t stare much. I expected considerably more attention in public.

4️⃣ Long pants work well. Aside from scraping against the metal rivets on my jean pockets. Shorts don’t work as well due to bunching and comfort issues, which is funny given how much marketing imagery involves outdoor & hiking use.

So what’s the verdict...

I’m convinced that most people will think recreational hip exos are neat for a few minutes but not something they'd use regularly or long-term.

Nonetheless, I think there are some valuable uses, potentially including certain older adults or others whose walking is limited by weakness/fatigue.

I’m also more convinced they’re often being marketed for the wrong things. The onus isn’t entirely on developers. Much of the nonsense comes from social media, influencers & even reputable media outlets.

Hip exos don’t need to be good at everything to be useful. We just need to get much better at identifying, and honestly communicating, what they’re actually good at.

Someone recently asked me if I wear a lab coat at work. Nope. 😁

The work we do involves testing devices in the lab, designing in CAD, and analyzing data.

It also involves A LOT of writing, conversations with collaborators, end-users and subject matter experts, and whiteboard sessions.

And sometimes exoskeleton R&D looks like a weekend walk around the neighborhood with a coffee in hand.

That’s because a big part of R&D in our field is connecting biomechanical principles and engineering designs to real-world use and impact.

Which means testing various types of wearable tech during all sorts of mundane activities — daily chores, work tasks, walking around the neighborhood, etc.

Learning what works biomechanically, what works practically, and where those two diverge.

One of the biggest mistakes I see from wearable tech developers is focusing too much on one of these, and not nearly enough on the other.

Always be testing. 🎯

#exoskeletons #exosuits #wearabletech

Why we dropped the batteries. 🔋

People sometimes ask me if we plan to add motors and batteries into the HeroWear exosuit.

Most are surprised to learn that we actually STARTED with motors, batteries, sensors, and even a phone app about 10 years ago.

And then got rid of them. All of them.

Our device evolution wasn't about adding complexity. It was about distilling the tech down to what was necessary, and what end users really wanted in the use cases we were targeting.

We didn't ditch the robotics and gadgetry because they weren't cool. They were!

Before HeroWear existed, our university prototypes and patents included a phone app, voice controls, and even an EMG armband to control assistance with different gestures.

Coolness factor was high. Engineers and media outlets loved the "advanced" tech and buzzworthiness!

But the harsh truth was that end-user enthusiasm was low.

Here's why we stripped out the complexity:

1️⃣ Customer feedback

Our earliest industrial customers were remarkably consistent. Several directly (and forcefully!) asked us to strip out the nice-to-haves and gadgetry.

They loved the back relief. But wanted something simpler, more reliable, and more durable.

2️⃣ Scalability

Even early on, customers were thinking about deploying hundreds or thousands of exosuits.

Charging batteries, maintaining robotics, managing an exo program, and overseeing data and cybersecurity all created barriers.

Today, powered industrial back exos cost ~10X more to deploy and sustain, depending on the device and use case. That makes widespread adoption and achieving ROI much harder.

3️⃣ Biomechanics

This realization came later, but was probably the final nail for us.

In our intended applications, passive elastic exosuits provided as much (or often more) musculoskeletal offloading than powered devices.

This surprises most people. But as the scientific literature on #exoskeletons grew, the evidence was clear.

It makes sense when you dig into the #biomechanics, timing of assistance, and weight distribution.

See the link below for more on the underlying physics and scientific evidence.

Anyway, that's been my journey. And it's shaped my design philosophy. Both in HeroWear and in other exploratory tech development projects at Vanderbilt.

There are many different applications for exos. Including or excluding motors isn't inherently better or worse.

But the complexity IS definitely better or worse for certain use cases.

I've spent the last decade learning a lot about the pros and cons, and honing my intuition for powered vs. passive, rigid vs. soft, and everything in between.

I'm more convinced than ever that exoskeletons and exosuits will become normalized in society.

At first, progress will seem gradual. Then it will seem fast. And eventually, it'll just be normal.

And a lot of people's health and well-being will benefit.

That's the future we're working toward.

And a little peek into how we arrived where we are today.

Cheers! 🍻

Exoskeletons, explained in layman’s terms:

🩺 Medical — helps people with injuries or impairments move

🏭 Industrial — makes physical work suck less

🥾 Recreational — portable leg swinger that makes walking feel easier 😃

#exoskeletons #exosuits #wearablerobotics

☝🏽 Has anyone here been using a consumer exoskeleton regularly for 3+ months? What do you use it for?

I’m curious to hear specifically from people who have owned and regularly used a consumer / recreational lower-body exoskeleton for at least a few months.

I’m talking direct-to-consumer hip, knee, or ankle exoskeletons rather than clinical/rehab or industrial exoskeletons.

If you’ve been using one for 3+ months, I’m especially curious:

• How often do you actually wear it now?

• What activities or tasks has it become genuinely useful for?

• Which device do you have, and how long have you had it?

• What did you initially expect to use it for, but found that you don’t?

• Are there situations where you deliberately leave it at home because the hassle outweighs the benefit?

❌ I’m less interested in first impressions, or in people using the device purely for research or because they’re already deep in the exoskeleton field and want to explore or stay abreast of the latest tech.

✅ I'm more interested in what long-term, real-world use looks like after the novelty wears off.

Hiking is relevant, but I’m also interested in mundane, everyday uses: walking around, commuting, errands, stairs, work, exercise, travel, chores, etc.

Reply in the comments or drop me a DM. Thanks!

#exoskeletons #wearablerobotics

Exo Impact = Physical Augmentation × Wear Time

It’s easy to over-optimize one and overlook the other.

Avoid these traps with industrial #exoskeletons and #exosuits 👇🏽

🤯 Biomechanics is mind-blowing.

An exosuit can relieve an estimated 26 MILLION pounds of cumulative peak force from a warehouse worker’s back each year.

We tend to get caught up thinking about and discussing EMG reductions, percentage offloading, or the biomechanics of a single lift.

But often we’re losing the forest for the trees.

Zoom out.

2,000 lifts/day
× 260 workdays/year
× 50 lbs of peak back force offloaded per lift
= 26,000,000 lbs/year

That’s equivalent in weight to 325 fully loaded semi-trucks that a person's low back doesn’t have to endure each year.

Now try to wrap your head around that cumulative physical relief over 5 years. 10 years. An entire career.

The biomechanics of a single lift matter. But ultimately, we care about what those small reductions add up to—for injury risk, fatigue, physical well-being, and people's ability to go home feeling better after years of physically demanding work.

#biomechanics #exoskeletons #ergonomics

🚨 Milestone: There are now 40+ peer-reviewed scientific publications on HeroWear exosuits.

So, at this point, I find myself less interested in scrutinizing every detail of every individual study. The more interesting question is: What have we actually learned from all of them?

Early on, I did pay close attention to every new paper that came out. I dove super deep into the methods and results. Sometimes I digitized the results from other labs so I could do more analysis, or followed up with the authors to learn more.

There simply weren’t that many data points—and I ate it all up.

But at some point, the work changed. The priority shifted.

Now, I often find the Discussion sections of individual papers to be too narrow, and not fully contextualized within what is currently a large evidence base spanning academia and industry. And I don’t think enough people are truly synthesizing and communicating what we’ve learned from all of it.

❌ The interesting question is no longer:

“What did the latest study find?”

✅ It becomes:

“What have we learned from ALL of them?”

Or:

• Where do independent researchers converge on the same conclusions?

• Where do lab findings hold up in the real world?

• Where do academic studies, field evaluations, case studies, and long-term experience point in the same direction?
• Where should we now have strong scientific conviction?

Here’s one place where I do:

The reductions in back loading that have been repeatedly measured in controlled lab studies really do translate to meaningful impacts in the real world:

Workers report less back discomfort and fatigue. This aligns with EMG and modeling studies. And across long-term deployments, we've seen real adoption and case studies measuring lower incidence of back injuries.

☝🏽 What’s most striking to me is that these real-world and long-term impacts have genuinely exceeded what we ever predicted from early university studies..

We knew that elastic exosuits could reduce loading on the low back muscles and discs, and could be practical and comfortable with the right softgoods design and features (e.g., on/off clutching).

What we couldn’t know from those early experiments was how much that offloading would matter after workers wore exosuits for years in the real world.

That’s why converging evidence matters. No single exoskeleton study carries that much weight.

But when evidence converges across researchers, methods, environments, and years?

That’s when a collection of studies starts becoming an evidence base.

And ultimately, I think that’s the bigger milestone here—not 40+ publications, but how much more we can now say with confidence than when this journey started.

Huge thanks to the global scientific community that continues to rigorously study, challenge, and advance this technology; and to all those using and advancing this tech in industry. Truly a team effort. 🙏🏽

#exoskeletons #ergonomics #biomechanics

🙏🏽 Feeling grateful to work on research, projects and technologies that aim to leave people and places a little better off. Don’t forget to occasionally zoom out and remember the big picture—what you’re working toward, and why. 🚀

#biomechanics #exoskeletons #wearabletech

💡 I learned a new term this week: possibility bias.

It’s when the fact that something could go wrong starts to overshadow the question of how likely it is to happen.

And I immediately thought: wow, this describes a fair number of topics in the exoskeleton field.

The pattern often goes something like this:

We can imagine a mechanism by which X could cause harm, therefore we treat X as meaningfully dangerous—even when there’s little or no evidence that this actually occurs (or is expected to occur).

Clearly, thinking through possible risks is important. Especially when using new or emerging tech like exoskeletons and exosuits that physically interact with people.

But there’s a big difference between:

“Could this theoretically happen?”

and

“How often does this actually happen? How large is the effect? And what does the evidence tell us about the real-world risk?”

I've had a number of private convos with researchers, industry professionals, and practitioners recently that have made me think possibility bias is more prevalent across our field than we'd like to admit.

I think this is a slightly uncomfortable conversation that we've shied away from in both academic and standards development communities.

As primary questions get answered, we naturally start probing secondary and tertiary questions. But sometimes that progression takes us toward increasingly remote or hypothetical problems.

Quite candidly, it’s something I've struggled with more and more through my own participation in standards committees. Sometimes it's difficult to know whether we're proactively addressing real-world problems or sometimes chasing theoretical ghosts built on hypothetical narratives.

And increasingly, I’ve come to believe there’s more of the latter than we acknowledge.

Standards have an important role in identifying hazards and protecting people. But they also need to mature alongside any emerging tech or field. I don't think we want them to be too proactive or too reactive.

Currently, I worry we're giving too much weight to plausible-but-unsubstantiated risks. In doing so, we can inadvertently spend large amounts of collective time and energy developing ways to address (or test for) problems that are exceedingly rare, or may not meaningfully exist, in practice.

Then assumptions can become embedded in standards or academic literature, where a theoretical concern can start to look like an established fact.

As exos continue to mature, I think we need to be more disciplined about distinguishing theoretical risks from demonstrated risks and making sure our response is proportional to the evidence.

Possibility ≠ probability.

Possibility bias is a useful new term (at least for me) and an under-discussed challenge for the #exoskeleton field as we distinguish the risks we can imagine from the risks the evidence actually supports.

I’m guessing there are some strong opinions on this one. Agree or disagree? Drop a comment or shoot me a DM.

🤔 This is pretty obscure, but I think I might be the world leader in wearing an exoskeleton or exosuit while driving to work.

And not just while driving.

I've worn various exos during work, play, weekend activities, meetings, hangouts, and much more—including, most recently, at my kid's rec sports game.

Several years ago, I embraced this mantra: Always Be Testing.

And it has served me, and the people I work with at Vanderbilt and HeroWear, extremely well.

Not because the exos assist with the driving itself, but because we learned so much about real-world comfort and usability during vehicle ingress, egress and usage, as well as range of motion, use in confined spaces, and compatibility when wearing or holding other items, and, more broadly, how the exo fits into the natural flow of daily life.

This has led to some real innovations and multiple patents, but also to some early realizations that the path we were on (or an early prototype we were testing) was actually dead on arrival. Often, it wasn't the biomechanical science; rather, the critical learnings came at the intersection of biomechanics, user needs, and user experience.

Living in the technology, not just testing it during structured studies, has led to some significant insights about design, research priorities, usability, and public perception.

Some of these are things I don't think I would have learned any other way.

If you work with exos or other wearable tech, I highly recommend this approach: learn passively by living in the wearable technology that you (and others) create.

Wear it. Live in it. Use it in public. See what you learn.

#biomechanics #wearables #exoskeletons

🚨 TN Maker Fest at Vanderbilt on September 26th 🚨

Check it out for an action-packed, family-friendly, one-day celebration of creativity and innovation at the Vanderbilt Wond’ry!

With over 60 exhibitors, from kid entrepreneurs and university labs to artisans and tech hobbyists, this festival is a showcase of passion, ingenuity, and community spirit.

Explore the latest in robotics, 3D printing, woodworking, and more across three exciting floors of hands-on fun.

Whether you’re a seasoned creator or simply curious, you’ll find endless inspiration around every corner. Free admission and free parking included!