Most people put a personal masseuse high on their wishlist if they were to become rich. What's not to like about a masseur on demand that knows your painpoints from too much time in front of a computer and can provide the panacea to the stresses of life?
Unfortunately, massages are expensive, and it can be hard to find a free appointment that fits into your daily schedule.
But now, Danish company Self.io has developed a massage robot that combines robotics, hardware components, machine learning, and computer vision to offer wellness massages that can help with stress, pain relief, and arthritis.
The company was founded in 2022 by CEO Julian Jørgensen and his brother Jimmy Alison Jørgensen, PhD, a professor of robotics. Self's massage robots target corporate offices, fitness centres and salons.
I spoke to Julian to learn more.
Why robot massage?
Robot massages help alleviate the chronic shortage of massage therapists in the industry. The robots can be available 24/7, addressing the issue of limited availability of human therapists. These would be amazing in airports or large hotel chains for people with jet lag and suffering from the pain of long-haul flights.
Further advanced scanning technology allows for highly personalised massages tailored to individual body types.
The idea came from Julian's previous experience as a software developer and entrepreneur.
Being 195cm tall, Julian was always susceptible to neck and back pain. He purchased the best massage chair he could find in Denmark, but it never provided sustained relief.
His brother Jimmy, a co-founder of Nordbo Robotics for the past eight years, had pioneered technology that seamlessly translated human craftsmanship into robotic precision.
Questions lingered about why the chair wasn't interconnected with the cloud, sharing and receiving data with its counterparts. Following a stroll past a tanning salon in Odense, the epicenter of robotics, the duo envisioned a future of massage robots tirelessly working 24/7, delivering high-quality massages at an accessible price point.
Julian explained that the massage offered "differs from traditional massage with hands. It delivers a percussion-style massage that is commonly used with athletes.
A physio onboards customers during the first session.
"In our experience, most people feel more comfortable with a human in the room for the first time just to give instructions and onboarding. So we use that opportunity to create a completely individual tailored and personalised massage program and emotionally onboard them. And that only takes about a minute or so.
Our therapists will personally onboard all users, and we have the capacity for 40-80 new massage programs in a day."
The customer lies down on a standard massage table. Clothing is optional. A camera detects a point cloud — a digital 3D representation of a physical object or space consisting of data points in three-dimensional space.
Self's system then offers personalised massages based on body scans to determine the appropriate massage movements.
The system uses data logging and machine learning to continuously improve the user experience. It collects biometric data, force measurements, and user interactions during each session. This information is used to personalise and enhance future massages for individual users.
The company has delivered about 5435 massages over around 1380 bodies, providing 800 point cloud data that we have.
"We can also use data like the ratings of the massage to feed into the AI to determine pressure. We can also use the data from the force sensor, remote control, and the NPS ratings.”
Julian recalls its popularity in fitness centres with 40-something women. Originally, they expected the robot to be popular with gym bros, but its first users were women 40 years old and over.
"We were pleased and surprised that this cohort had so much trust in this new tech."
A new approach on product testing
Self is not the first massage robot on the market, but the startup took a very different approach than its French and American counterparts.
The company developed a secure, stable, and high-quality robot on a budget of less than $250,00. This encompassed a spectrum of components: data infrastructure, web interfaces, payment systems, user management, monitoring, alerts, hardware, sensors, AI, machine learning, compliance, patenting, freedom-to-operate reports, cameras, salaries, legal fees, consultants, and much more.
Further, it developed the robot on the market.
"That required a lot of trust from our customers. And we also needed to trust the product. But there was no bad outcomes. There were, of course, some cases where the robot wouldn't start or it stopped too early in the first year.
But the strategy was to test on the market because it gets real feedback, and we build what people really want."
When initially deployed in Norway at Europe's largest massage chain for user feedback, 75 per cent of 100 users said it was worth the money and gave it between seven and 10 stars. Since then the company has gone on to refine the product further,
Julian notes:
"People are reluctant to invest in buying the entire robot asset because it's quite a lot of money. But renting it on a monthly basis is a bit less intimidating."
The robots are rented out for €2000 per month for 600 massages to corporate offices and fitness centres under a B2C subscription model.
Customers in a fitness centre can receive a massage for about €8-€25 per month depending on the membership type. This makes massage accessible to all.
"Once you have the CapEx in place,the expenditures of the robots, then it's just the rent or the electricity needed."
Self’s CapEx now only costs about €20.000.
How does the Self robot differ to what else is on the market?
Robot massage technology has been gaining traction in both France and the United States.
In France, Capsix Robotics offers a full-body AI massage bed with two robotic arms and a digital screen, while Aescape in the US provides a friction-style massage which requires the user to wear a suit because,
otherwise, it could create so much friction that it can give burn marks.
Aescape has raised $85 million and partners with luxury fitness brands like Equinox in the US, while both Capsix and Self aim to make massage more accessible to the masses.
Speaking of friction burns, what about safety?
I was curious about safety, can a robot go wild?
Julian notes, "We were super surprised that people they'd never asked about safety. We pitched investors, customers, and users. But none of them really asked about safety. They just assumed that it was safe."
The massage is offered a pressure scale. But the robot cannot move faster than it can hurt a human. And the hardware has so many safety mechanisms- the hardware and firmware — filled into the arm.
There are also hardware restrictions on how hard it can apply pressure. It is CE-marked and very safe to use.
And its just the beginning, Julian explained:
"In the long run, our vision is to develop a sensor-embedded glove that can train the robot to replicate various massages."
The company aims to refine its tech to include autonomous robots based on full body scans and scale its services to corporate offices.
With advancements in AI and robotics, we can expect to see even more sophisticated and versatile massage robots in the future, alleviating stress and improving overall well-being.
Lead image: Self. Photo: uncredited.
Would you like to write the first comment?
Login to post comments