Shaping up: Cyberdyne's exoskeletons are the firms to meet international safety standards
Within the last year companies have started to demonstrate the first robotic exoskeletons – wearable robots – for industrial and medical uses. The first applications that bring people and machines closer together than ever before are appearing, and a new generation of benefits is being realised.
In the medical sector, exoskeletons for rehabilitation and even long-term personal use at home are receiving approval. Companies such as Israel’s Rewalk and Ekso Bionics of the US are giving people who suffer from spinal injuries the opportunity to walk again.
The first systems have their roots in the defence sector. In the US, Lockheed Martin has demonstrated powered and unpowered exoskeletons. Developed by Ekso Bionics and licensed to Lockheed, the Human Universal Load Carrier (Hulc) is being tested by the US military. The battery-powered, hydraulically actuated “lower extremity” suit enables soldiers to carry loads of up to 90kg for up to 20km on a single charge.
Hulc’s development has been behind closed doors. A lighter and more efficient version of the Hulc, which can also assist soldiers in lifting loads, was reportedly last seen three years ago at a trade show. In contrast, Lockheed has been keen to publicise its unpowered exoskeleton for industrial uses. The Fortis increases an operator’s strength and endurance by transferring the weight of heavy loads from the body directly to the ground.
Last August the US Navy purchased the first two suits to evaluate their use in shipyards. Adam Miller, director of new initiatives at Lockheed Martin Missiles and Fire Control, says: “Ship maintenance often requires the use of heavy tools, such as grinders, riveters or sand blasters, which take a toll on operators. Wearing the Fortis, operators can hold the weight of those heavy tools for extended periods with reduced fatigue.”
Outer layer: Powered suits such as Lockheed Martin's Fortis are being developed in the US
It isn’t just the US military that sees the potential of exoskeletons in industrial settings. French company RB3D has been developing exoskeletons and cobots for industrial use since 2003. The company’s Hercule exoskeleton was developed first for defence applications and is at the prototype stage. However, the company’s range of collaborative robots, which shares technology with the exoskeleton, can be purchased today.
What makes a collaborative robot, a cobot, different from a standard industrial robot is that it works next to a person, together with them, bringing more strength and capacity to a process. The cobot does not require the equipment and guards normally used in a robotic cell.
RB3D has patents in key areas to ensure safety, says Olivier Baudet, business manager at RB3D: “The patents are in the architecture of the hardware to make sure there is a high level of redundancy in the control systems. They align with the latest international rules and regulations so they can be used without risk. It works by knowing the position of the human operator constantly from the position of the hands.”
Another key difference is that a cobot isn’t autonomous and doesn’t need to be programmed for one specific task. Instead it uses a combination of human operation and high-level software, so it can be reused for different tasks. Baudet says: “We are more expensive now than industrial robots from the likes of Fanuc, Kuka or ABB. Around three times the investment compared to a £30,000 robot. But we don’t need the surrounding equipment, a cobot doesn’t have to be programmed and can be used for different tasks. The final cost of a robotic cell is higher.”
However, Baudet says that cobots should not be seen as superior to robotic cells. They are, he says, complementary. “Where there is a fixed process there is a clear return on investment for reducing the number of workers. In these cases a robotic cell is the best solution,” he says.
So far, the company has installed 15 machines in various applications. The typical scenario for the use of cobots is the grinding of metal parts in foundries and finishing workshops, says Baudet. The company is seeking to build sales in this area when it starts exporting the cobots later this year. “We are focusing on the most strenuous tasks first,” he says.
“The most challenging barrier is reducing the cost of cobots and proving their value through the increase in production rates and capacity they enable. They improve health and safety and empower people. With a cobot, weight and power are no longer issues, and risk is actually reduced.”
Sales of the cobot range will help RB3D to further develop its Hercule exoskeleton for industrial uses. A key difference between Hercule and medical exoskeletons is the element of control. Medical exoskeletons available today are programmed to work like a robot for people who can’t move their legs, says Baudet. Hercule is designed to follow a person’s movement in order to augment it, like the cobots. It does not follow a pre-programmed routine.
“We see two key markets for the first industrial exoskeletons,” says Baudet. “Strenuous tasks with manual tools that require a lot of torque and the movement of heavy loads for logistics. In these situations we’ve shown that the use of an exoskeleton makes the work a lot less demanding.
“The industrial and logistics market is the biggest opportunity. But the first market is medical. Military applications will take time because the specification is so tough.”
Professional partnership: Cobots offer more strength and safety in some industrial processes
The immediate challenge is power. Hercule is able to operate for four hours from lithium-ion batteries in normal conditions, thanks, says Baudet, to the high power-to-weight ratio of its mechatronics. Longer usage times will be made possible with improvements to its exoskeleton and an increase in the power density of batteries. The power problem can be minimised by making the batteries easy to swap and charge.
Another design challenge has been to make the suit fit a variety of body shapes so it can be worn and removed quickly. Hercule achieves this in under a minute, says Baudet: “An exoskeleton should be like putting on a jacket and be adjustable. Hercule is designed to fit 90% of people.”
Meanwhile in Japan, the ageing profile of the population has encouraged R&D into exoskeletons, as companies look into ways of keeping people working for longer. Activelink, a subsidiary of Panasonic, has leased out the first test models of its Power Loader device to agricultural and logistics companies. It plans to have 1,000 of the devices leased out by the autumn.
Baudet says that Europe is not trailing behind Japan in the field of exoskeletons. “Japan has started early with medical applications, but we’re not following them with industrial applications,” he says. “As far as I know, Cyberdyne’s exoskeleton takes a long time to get in and out of and isn’t ready yet for everyday use. Panasonic’s is still at trial stage.”
The ominously named Cyberdyne, a spin-off from the University of Tsukuba, is also making progress in Japan and looking to develop international markets. The company’s Hal – Hybrid Assistive Limb – became the first device in the world to be ISO certified as a wearable robot for workers and care givers last November. The lower limb version of Hal wraps around the user’s lower back, hips and thighs and is powered by a lithium-ion battery. A myoelectric sensor detects spinal activity to operate actuators on the hips to provide additional support.
Demonstrations have shown the unit assisting a person in lifting two 20kg weights. The device has also been tested by Japanese construction company Obayashi.
It remains though that exoskeletons available for both medical and industrial applications today are bulky – made from metal, wires, circuit boards and motors and linkages. Although impressive feats of technology, they are a realisation of a sci-fi inspired vision from the 20th century.
A group of researchers in the UK is looking to the next stage of exoskeletons, beyond robotic suits that move mechanically. The future, according to the Wearable Assistive Materials (WAM) research project, is exoskeletons that are worn like clothes or bandages. These exoskeletons use smart materials to mould to the shape of your body and chemical processes and magnetic gels to mimic natural processes in skeletons and muscles.
The £1 million WAM project is taking the first steps in developing the materials and technologies needed to make the vision of a wearable exoskeleton a reality. The project is encompassing a broad range of areas, from orthopaedics and biomechanics to new sensor technology and novel 3D printing techniques.
One of the first tasks of the project has been to assess if there is a demand in the clinical and patient world for exoskeletons in the first place. Clinicians and specialists in orthopaedics involved in the project have run focus groups to assess patient needs.
Peter Smitham, a clinical lecturer and orthopaedic registrar at University College London, is a WAM researcher. He says there are other factors driving the recent uptake in exoskeletons, in addition to the technological aspects. “Exoskeletons are beginning to be included in litigation costs to aid rehabilitation. There’s also a build-up of publicity and interest that is steadily feeding on itself,” he says.
“But they are still prohibitively expensive. And no one seems to have asked patients what they want. The machines take a long time to put on and take off, move at about a third of the speed of walking, and have limited battery power. People acclimatise to using a wheelchair and don’t want to be slowed down. They also need to be packaged easily, so they can be taken in and out of cars quickly.”
There are also concerns around the complications that may arise when people with spinal injuries survive for longer and are placed in new positions by exoskeletons, says Smitham. For example, people with spinal injuries often can’t feel pressure sores. There is also a rapid loss of bone density in people with long-term spinal injuries.
Dr Catherine Holloway is a lecturer in accessibility engineering at University College London and also works on the WAM project. She says: “Exoskeletons are going through the classic new technology curve. They’re bulky and unwieldy like the first mobile phones or computers were. It will be the second or third generation before they are adopted in the mainstream.
“That’s not to take away from what the exoskeleton companies are achieving.”
Holloway says there are two main approaches to exoskeletons: the traditional, wearable robots on the market now, and the emerging area of soft robotics. The WAM project is looking at using smart materials and soft robotics.
“As you flex your muscle, the material has to adapt to the change in volume and give appropriate support, either additional or less,” she says. “Robotic solutions are driven by the angle of your leg, using the control systems. Wearable assistive materials will adapt as needed depending on the scale of support to be provided using either a variable control or an on/off approach with chain mail.”
Researchers first looked at the possibility of using ceramic tiles as the main support structure. This was changed to a 3D printed pattern that closely resembles chain mail, says Holloway. A wetsuit, chain mail material allows for more flexibility. Pressure, heat and other types of sensors within the material provide feedback to the control and actuation system.
The movement of the material is being achieved by mimicking the chemical processes and patterns that occur in actual muscles. Smitham says: “Instead of hinges and actuators, we are using magnetic gels and chemicals with in-built flexibility.”
The team is using vanadium oxide, which flexes when exposed to an electrostatic potential and exhibits up to 10 times more strength than skeletal muscle, as the chemical actuator, and magnetic ferrous gels as the dynamic controllable mechanism.
The output of the WAM project will be a working prototype of a material that can bend and shape itself and lock into position to provide support when required by the wearer. Holloway says: “Once we have a working example with all our desired properties, the other engineering challenges can be addressed, such as the power supply and the fibres within the magnetic gel.”
The project has grown to have lots of layers of research – from nozzles for 3D printers to the development of a fully instrumented skeleton equipped with sensors to test new materials. This has required the development of new silicon-based sensors to measure and flex in the ferrous gels.
There is potential for spin-off opportunities in the development work and materials from the WAM project, believes Holloway, who adds that the rate of progress in assistive materials depends on the level of investment the research receives.
“Traditionally, assistive technologies don’t get a huge level of investment,”
she says. “At the moment these materials are expensive and difficult to get hold of. But there will be a critical-mass moment when the materials become cheaper and more available. Then they will be used in applications that today seem wild and strange.”