Versus Fitness: Developing A Smart Gym

VersusOver a year ago, I wrote a post about developing with the Kinect and how I was working on a project that revolved around it. Fast forward to today, the project is now an officially launched gym that is also known as Versus Fitness.

What is Versus Fitness?

Versus is a system that has gamified fitness. By utilising different sensors and technologies, it is able to measure 3D motion, pressure, force, acceleration and power output of over 200 different gym exercises (and counting). With each proper repetition (or rep) that is executed, the user not only gets the rep counted by the system, a score is given based on the above measured parameters, and the score is scaled based on the user’s weight and height. In that way, 2 people of different weight and height doing the same workout can compete against each other on almost equal terms (Hence the name Versus). In case the term “wearable technology” comes to mind, no, there is nothing that the users need to wear to get their exercise tracked (maybe except a heart rate monitor, but that is purely optional). Just check out the video below.

How I got involved?

I started working on the Versus Fitness gym since late 2013 and it was purely by coincidence. Someone who knew Brad Bond (the founder of Versus Fitness) was at the RMIT Sports Engineering Lab on one of those Uni open days and he saw a novel sensor technology that would suit Versus. After a series of meetings and discussions, a research contract was set up to further develop that technology for Versus. This was partly funded by the Victorian Technology Development Voucher. At the same time, they were also looking for an additional team member to work on motion tracking algorithms. That’s where I came into the picture. Long story short I was offered a contract role on the Versus project which was partly funded by the Enterprise Connect – Researchers in Business grant (this has been replaced by the Entrepreneurs Infrastructure Programme). Kudos to Aaron Belbasis who was a key connector/initiator who brought everyone together and who was also one of the key researcher who helped develop the novel sensor tech. There’s a bit more details about the RMIT-Versus collaboration here.

What Tech are we talking about here?

One of the sensor technologies came from the research collaboration mentioned earlier. The team at RMIT calls it a “sensor-less sensing platform”. The closest thing would be Force Sensitive Resistors (FSR) like the ones from Tekscan. If you had a proper look at the video above, you will see the “sensor-less sensing platform” used in the floor exercises and some of the running exercises. Basically its a sensor that measures pressure.

There are other sensors that were developed or customised for tracking motion and a number of them are available off the shelf or at least purchasable online. In fact some of the sensors (like load cells and accelerometers) are similar ones typically used in the manufacturing, or automotive industry. A lot of custom fittings, enclosures and mechanisms were designed for the sensors before they could be installed in the gym. Majority of the design were done in-house and prototyped with the help of a MakerBot replicator.

But what really made the sensors (tracking system) worked effectively are the smart algorithms that processes all the sensor data and accurately identifies when each person is performing the exercise properly and evaluates how well he/she has done it. Initially when designing the algorithms for tracking each type of exercise, it all seemed pretty straightforward; but as things progressed, it turned out there were quite a few more considerations – e.g. filtering out “incorrect” movement data that resembled an actual rep, or profiling movement data from users of different abilities (or fitness level) etc.

Perception & Reality

Another important part of the system is the “gaming interface” or the “gaming control centre”. It is the personal trainer’s assistant. It relays to the users what exercises to do, records their performance, stores the performance data in a database, reminds the user how well they did previously (their Personal Best), manages the equipment (to some extent), and ensures that every exercise station is in sync so that the workout runs smoothly. That allows trainers to focus on one of the things they do best: scream at motivate people.

So with the combination of the sensors, smart algorithms and the gaming interface, this means: real-time tracking, with feedback of the users’ performance (score) or technique delivered right after each completed rep, and an overall quantified workout so users know how well they fare compared to their previous workouts (and with other users).

Future Developments?

The very first Versus Fitness gym is based in Moorabbin and that has seven different exercise stations (as seen in the video above). One could call that the full Versus experience. There are a number of possible developments in the pipeline. One is the development of new exercise stations to increase the type of exercises that can be tracked. Also, there are possible opportunities to customise the system for the elite or professional athletes, or even rehabilitation applications. Something that is definitely in the works is a “multi-station” concept – a single exercise station that has several sensor solutions allowing tracking of a few different types of exercises (e.g. dumbbell, kettlebell & floor exercises). This significantly reduces the footprint of the equipment and would suit small gym spaces. In fact this is currently on trial in a gym somewhere in Australia, and depending on how things go, you might start finding the VS logo in many more places!

Laser Additive Manufacturing – applications in medical technologies and beyond

The Advanced Manufacturing Cooperative Research Centre (AMCRC) organised a workshop last week on additive manufacturing and its impact on medical technologies. Organised in conjunction with RMIT and Bio21 Cluster, the aim was to promote the new technologies and also educate the attendees on the network and resources available to help businesses adopt them.

The latest in laser additive technology includes Selective Laser Melting (SLM), Electron Beam Melting, Laser Metal Deposition, and Laser Sintering.  Their main advantage is the ability to build complex shaped objects using biocompatible metals such as Cobalt-Chrome, Titanium, and 17-4 PH Steel. This then simplifies and quickens the process of customising orthopaedic and dental implants; and using additive manufacturing basically means less wastage of materials compared to traditional subtractive methods.

Anatomics, who had a rep presenting at the workshop, is one of the companies applying this technology in the medical field. They are a Melbourne based company, specialising in cranial and maxillofacial custom implants. They also produce BioModels based on CT scans or MRI, which allows surgeons to have a better visual and feel while diagnosing and subsequently help improve surgery planning.

The potential for laser additive manufacturing is huge, but currently most of the ‘action’ are still in the universities, research institutions, and hospitals. That is where government funding programs and organisations like the AMCRC come in to help bring additive technology into the industry or even to form startups. The challenge is that the size of the local market (Australia) is too small for this technology, so it must definitely go regional or even global for a commercial entity to be viable. Then even before that, to build up an environment that embraces entrepreneurship, innovation and collaborative efforts.

Looking at Kickstarter – the latest trend in global crowd-source funding, if you search “3D printer”, there are at least 10 projects trying to come up with their own machine for 3D printing (which is the more common name for additive manufacturing). They are typically motivated by three reasons:

  1. Existing 3D printers in the market are too bloody expensive
  2. They would like certain features or capabilities missing in existing printers
  3. Read point 1 again.

Although currently these Kickstarter projects are only suitable for printing polymers and objects smaller in dimensions (around 100x100x150mm), but their low cost of entry (between USD$700 – 2500) is rapidly promoting the use of additive technology and they just might be the tipping point for design and additive manufacturing. Just check out this project that is already getting close to 20 times its original funding goal!

Finally I found this Ted talk that gave a very good overview of additive manufacturing from a designer’s point of view: