The challenges of making Smart Sports Garments

‘What is a Smart sports garment?

Smart sports garments or smart performance garments is a relatively new product segment in the consumer sports tech market. There are probably different views of what the definition should be, but for the purpose of this post, it is a sports garment with embedded sensors/electronics. The main functions of sports garments include providing covering, protection, comfort, ease of movement and some might say making the athlete more aesthetically pleasing. Then with the added sensors and electronics, there generally are two different types of secondary functions.

The more common one is the passive function where sensors monitor stuff on an athlete, either physiological measurements or physical movements. It can make smart evaluations based on the data and give real-time feedback suggesting to the athlete that they should push harder or rest or correct their technique etc. But the decision to act on that suggestion still lies with the athlete or coach. There is also the not-so-common active function where the garment does something to the user. For example giving electrical muscle stimulations or possibly electric shocks. But so far the “electric shock” feature is only found on a wristband and hasn’t extended to any other wearables yet. I am not sure why that is the case. For EMS, it has been said that it helps with muscle strengthening which is good for rehab or as a complementary training tool. But I will not go into it since it’s beyond my area of expertise.

R&D in Melbourne

A while ago, I had the opportunity to be a lab rat for a mate’s PhD thesis. He has developed a patented novel technology to measure muscle activity and hopefully able to predict the risk of muscle and knee injuries in elite athletes. The experiment I took part in was basically collecting a bunch of data from this novel sensing technology, wireless electromyography (EMG) sensors, a motion capture system, and a bike trainer. Unfortunately, it also involved me pedalling for my life.

How is this relevant to smart garments? Well, the novel sensors and EMG sensors were all hidden under a compression garment with motion capture markers secured on the outside. The compression tights ensure that the sensors remain where they are (and reliably capture data) and they also (coincidentally) facilitate motion capture. Albeit it was a very crude way of combining the sensors and the 2XU tights, it was a functional prototype (of sorts), and the ultimate goal would be to have those novel sensors built into compression tights.

ppuy6097

Lab rat in action

As we discussed further on commercialising this novel sensing technology for smart sports garments or developing smart compression garments with any wireless sensors, it became apparent that there are a number of challenges. Here’s just a few:

Washing and durability :: A sports garment is going to get sweaty and smelly a lot more than everyday garments. So it definitely needs to get washed. Most smart garments in the market have an electronics module (IMU, BLE module, battery etc) that is removable because they will not survive a tumble in the washing machine. However, there are still conductive pads or conductive yarns (for electrical connections). Would long term washing affect their conductivity and so usefulness?  (A research has shown that most conductive threads will be affected although some hold up better.)

Sensor data accuracy :: In order to capture accurate & robust data, the sensors have to be positioned in the correct location each and every time the smart garment is put on. For measuring stuff like heart rate or EMG, it needs to maintain skin contact for proper measurements. If sensor positions are off (by a bit too much) or skin contact is not maintained, the data collected becomes meaningless and cannot be compared with previous data sets. Not to mention the effect of sweat on EMG electrodes.

Custom fitting :: This relates closely to the above point. Most sports compression wear are made in standard sizes. Sometimes one might find their compression garment being a bit too long at the legs or too short for the arms or too tight around a joint and too loose at a certain spot. It’s fine on a regular compression garment. But when sensors come into play, especially when there is fabric type of sensors (that measures compression or stretch), perhaps a custom-fit garment could be a more optimal solution.

Application :: This is possibly the most important challenge – designing a smart sports garment that solves a real need. It could be a very niche area or a wide-spread problem. But the starting point would be talking to athletes, coaches and sports scientists, to identify where the need is or what needs to be tracked. Then the smart garment that is developed would be a solution and not just a cool piece of technology.

What’s in the marketplace

Having said that, over the last 4-5 years, more than a handful of companies have taken up these challenges and developed their own smart sports garments. A quick search on google shows that there are at least 5-6 smart sports garments in the market.

Brands / Companies
Measured parameters
Heart rate Breathing frequency EMG Motion 3D motion (joints)
OmSignal
Hexoskin
Athos
Myontec
Heddoko

cvetfcjxgaa1ngv       hexoskin-6_v2

OmSignal and Hexoskin have smart garments that are an extension of heart rate monitors with an added IMU (Inertia measurement unit) which provides parameters such as breathing rhythm, running cadence, step count and more. While they both seem to be generic fitness trackers when they first came out, it looks like Omsignal has now dropped their original Omshirt and focused on a women-specific product (the Ombra) for running. This might have to do with a review like this: link.

       allsport_kit_600x680_01   1-ncipvwdjblutc7zvxalxaa

Myontec and Athos are smart compression garments with surface EMG sensors. The point of putting on these garments is for the user to know what’s going on with specific muscle groups during their run, cycle or gym workout. Myontec is focused on the lower body (quadriceps and hamstring) with an emphasis on running and biking, while Athos covers the whole body looking at general strength training. It is cool that their accompanying software/app provides feedback of which muscles should be activated more during a squat (or other exercises) but I think it might be better if they could correct a user’s posture/technique that is causing the wrong muscles to be activated.

aaeaaqaaaaaaaaymaaaajdflodaxnjuxltgwmzetngzmmc05owm3lti3mzmzythjowywmg

Heddoko is a full body compression suit that measures a user’s 3D motion much like the Xsens suit. The difference is that the Heddoko suit uses less number of IMU and has embedded stretch sensors, which makes it unique. Assuming the measurements are accurate and repeatable, it has lots of potential applications in sports biomechanics and injury prevention. But based on this video, they are still validating their sensors and trying to work out specific applications.

Some additional thoughts

On one hand, it is cool that there is all these performance tracking technology available to the average athlete – such as wireless EMG and 3D motion analysis (again, assuming the measurements are robust). On the other hand, I wonder if the benefits would outweigh the costs because they are mostly quite expensive and I am not sure if the average gym goer would need that much information about their workout. Perhaps they would be more useful to elite or professional athletes, especially where professional teams have coaches and sports scientists to analyse the data, and give custom feedback. They could also couple it with video playback and analysis so that there is more context to the data.

I think for the average athlete, a smart garment might be useful if they are going through physical rehab and need to monitor certain movements or muscle groups while under the guidance of a physical therapist. Or if they are trying to pick up a specific skill like throwing a football or baseball (In fact, there are sensor embedded sleeves that do just that, which I might discuss another time). Basically, there should really be a specific ‘pain’ to solve. A smart garment with a generic health and fitness application is probably not going to be of much use. Wristbands and smart watches already try to do that.

Do you already own a smart sports garment or are thinking of getting one? If yes, do leave a comment. I would love to hear your thoughts and what you would use it for. Thanks for reading!

Advertisements

4 thoughts on “The challenges of making Smart Sports Garments

  1. David Lloyd

    I read with interest your blog on smart sports garments. My comments are based on having conducted research for over 20 years (https://scholar.google.com.au/citations?user=FisilIUAAAAJ&hl=en) into neuromusculoskeletal biomechanics of the causes and way to prevent sports injuries, particularly the knee. In this we have also developed sophisticated neuromusculoskeletal computer models, which has used data from high end laboratory motion capture (forces plates, EMG, optical motion capture, e.g. Vicon) to drive these models. I am now firm believer of moving from the laboratory to the real world using multiple wearable sensors, be they discrete sensors (IMU, EMG etc) and/or garments with embedded sensors/electronics. However, it is difficult for the individual, and even the hardened scientist, to assess all the multiple sensor information: this needs to “condensed” to the “crucial information” relating to tissue injury and prevention.

    We must not forget that through many, many studies we have gained much scientific and engineering knowledge on what constitutes the crucial information on the causes and ways to prevent musculoskeletal tissues injury. Some of these have been successfully translated to real-world training programs proven to lower injury rates, i.e. the knee and ACL. Furthermore, the literature is deep with scientifically developed and validated neuromusculoskeletal biomechanical models of, for example the knee, and we have a firm understanding that accurate estimation of tissue loading requires many neuromusculoskeletal factors to be incorporated into the models, and that these need to be subject specific. Furthermore, this literature is not necessarily reviewed by the wearable sensor/device world, and this knowledge and knowhow must be maintained and translated from the laboratory instrumentation and environment to the real world wearable instrumentation and environment.

    I read with interest the “patented novel technology”, which is a step forward for integration of multiple wearable sensors to provide the crucial information of ACL loading. However, it appears this may have been developed without full regard to vast literature on developing and validating neuromusculoskeletal models – these models have been validated against actual measured forces and strains in tissue (see the Grand Challenge Competition to Predict In Vivo Knee Loads). It is well known from literature (biomechanical models and cadaveric testing of the knee) that the ACL loading is due to 6 dimensional external loading, not just muscle forces, and certainly not just the quadriceps and hamstrings, e.g. gastrocnemi. Furthermore, all knee muscles have the potential to both load and unload the ACL. Also estimating muscle forces from EMG is not as simple as proposed – one needs more complex musculotendon models. The anatomy of the knee, lines of action of ligaments and slope of tibial plateau are different for each person, and have been show to dramatically effect ACL loading and its estimation from models – one model does not fit all. Nevertheless, the “patented novel technology” is a step in the right direction, but it is clear we need to translate the excellent and long standing research into these new wearable devices.

    In summary, we need to move what we have learnt from laboratory based studies to the real-world, and the development of wearable devices that mimic the laboratory technologies and findings will certainly help in preventing and rehabilitating neuromusculoskeletal injuries.

    Liked by 1 person

    Reply
    1. Julian Chua Post author

      Hi David, thanks for your insightful comments! I will definitely have a look through the references you mentioned. Would love to collaborate if there is ever an opportunity.

      Like

      Reply
  2. Pingback: Applied Sports Science newsletter – January 21, 2017 | Sports.BradStenger.com

  3. Pingback: Sports Science: Week in Review, Jan 9-15 | Sports.BradStenger.com

Leave a comment, share your thoughts. If your a SPAMMER or building back links to unrelated sites - this is not the site you are looking for!

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s