Re-imagining an Industrial Robot

‘Imagine a robot.’

‘What is the first image that comes to your mind?’

I asked this question to a bunch of my geeky friends while watching a sci-fi movie. The answers ranged from C3PO in Star Wars to the evil machine Ultron, from Avengers. For me, however, the ultimate robot has always been Vicky, the little girl from the quirky family in the popular show ‘Small Wonder’. She was, what I believed, the ideal robot –an efficient machine that was a natural part of human lives, and at rare moments, displayed reactions that can be qualified as emotional.

Hold that image in your mind. Now, let’s take a step from science fiction into the realms of reality.

Imagine a robot you have seen in action on a factory floor. Probably a robot in a warehouse or in a car manufacturing facility.

The image above is of a bunch of robots working in unison. Quite a contrast from our sci-fi imagination, isn’t it? Unlike the uber-sleek entities in fiction that interact fluidly with humans, real-life clunky intimidating metal-bots are still not designed for human proximity. ‘What about ASIMO?’ one of my friends argued, ‘He seems to bring robot-reality closer to science fiction. He is sleek, and he is so smart!’ Humanoid bots today have helped us conjure a future where robots can be just like people- have a mind of their own. But despite all the romanticism for robots, we are still far away from a world where robots are akin to humans and can replace the humans as a superior race. But that doesn’t stop robots from becoming increasingly important in the digitised industrialization era as they replace, complement, and augment human effort at affordable costs.

Basically, robots can’t think yet – we create them and hence infuse our logic in them. They are simply doing the job they are assigned to do using our logic. But the fact that robots do not have will power (as yet) makes it even more important and morally imperative for us to create ones that would not harm us while only trying to help. In an unfortunate incident in July, 2015 a 22-year old man was grabbed by an assembly robot and crushed against a metal plate while working in a Volkswagen plant in Germany. In science fiction this would have marked the rise of the machines but in reality, the horrific incident was unsurprisingly due to a human error.  Nevertheless, this incident has delayed the ubiquity of the robot on the factory floor, and eventually the human home. The need to create robots that are safe to work with was getting stronger and stronger.

Another roadblock for the robot to become ubiquitous is the high cost. My friend, the Curious Joe, googled ‘ASIMO cost for renting’ and gasped at the search result. Two museums in Japan had apparently rented the ASIMO for $150,000 a month! ‘Industrial robots are, of course, cheaper,’ we thought.  But is it affordable enough? An Industrial robot can cost around $50,000 to $80,000, and with application-specific peripherals added, the cost can be around $100,000 to $150,000. Take into account the fact that a lot of manufacturing happens in developing countries where cost of labour is still extremely low, it looks difficult for robots to make a case for themselves.  

Enter soft robotics.

As the name suggests, these are robots made of light-weight soft materials, mostly polymers, that work without motors and actuators. Doesn’t seem legit? These robots use pneumatic mechanisms – they use air pressure for movement, rather than your traditional motors and actuators. Soft robotics is a novel attempt to enhance mechanical automation through flexible materials. It is not surprising, therefore, that George Whitesides, a professor of chemistry at Harvard University is one of the flagbearer-researchers of soft robotics!

Most soft robots are bio-inspired. They try to mimic the greatest machines ever built – living bodies. A good example of a bio-actuator is a muscle. It helps perform dynamic tasks in uncertain circumstances. Soft robots with their flexibility, adaptiveness, and polymer-body are very close to mimicking the living body and therefore replacing some of the complex activities that traditional robots would find hard to perform.

 ‘A polymer body with no traditional actuators would drastically reduce the cost,’ one friend remarked. In a bid to find out more, I researched on DIY kits for soft robots, and techniques to 3D print basic soft robots using basic polymer materials. Some of these robots can be built for as low as $20. I am not comparing a $20 robot to a $150,000 a month ASIMO in its abilities. But the bottom-line is that soft robotics has cost disruptions to sit up and take notice.

The key points that make a compelling case for soft robots are:

1.      They are dextrous, like humans

2.      They are easily manoeuvrable, akin to a swarm of bees or birds

3.      They can work closely with humans without creating risk of injury or death

4.      They are drastically lower in cost compared to traditional robots

If soft robots are so disruptive to traditional robotics, why haven’t we seen more of them yet?

Soft robots are often harder to control due to the ‘spine-less’ structure and lack of actuators. Speed, again, is an area where the soft robot often scores less than the hard counter part due to the soft robot’s pneumatic actuation method. Also, large scale field reliability is yet to be tested. There are start-ups, however, that are leading the way in bringing soft robots to the mainstream. A Harvard start-up called Soft Robotics is building end-of-arm tools or grippers for manufacturing, warehousing, and food processing sector.

Soft materials and sensors have been growing in interest for researchers but soft robotics got the much needed momentum in 2013 when EU established Robosoft - a co-ordination action for soft robotics. Since these researchers come from various backgrounds, Robosoft provided the much needed platform and framework for the field to develop through research co-ordination among 22 institutions globally. Since 2013, we have seen breakthroughs in soft robotics – soft artificial skin, soft octopus, and very recently, soft actuators.

Technology leaders of the Tata group dwelled on soft robotics in early 2015, during a Tata delegation visit to Harvard university. CTOs, and technology leaders of several Tata companies met with professors at Harvard University and the Wyss Institute of Biologically Inspired Robotics. Some very interesting work was showcased to the delegation:

1.      RoboBees: Dr Robert Wood, founder of the Harvard Micro-robotics lab, showcased soft light-weight robotic bees built with multi-material. These robots mimic the bee and are important experiments to strengthen mechanical intelligence. Mechanical intelligence is a term coined by American psychologist and author Edward Lee Thorndike who defines it as the ability to effectively control one’s body and manipulate objects. Therefore, the ability to use feedback mechanisms to reduce controller capacity and create dynamic responses would help accomplish complex jobs without human intervention.

2.      Soft exo-suit: Dr Conor Walsh, Founder of the Harvard Biodesign Lab, built a soft robotic exo-suit that mimics the biomechanics of walking and using soft sensors, functional textiles and advanced controls, significantly assists walking and working with loads. Further work in this domain could help create exo-suits that work with humans to reduce load and increase efficiency while retaining the advantages human dexterity offers.

3.      Elastomeric grippers and tentacles: Dr George Whitesides’ lab at Harvard has been working to create soft robots that can perform complex motions using ‘intuition’ rather than through complex programming. This can therefore be useful in handling fragile objects and unevenly shaped objects, a challenge currently in logistics as well as assembly lines.

How can Tata play a role in the unfolding of the soft robotics story?

Tata group collectively has what very few companies around the world can provide – a big playground for soft robotics on the factory floor. From automobile to watches, from steel to chemicals, we have factories in diverse domains that could potentially benefit in cost, automation, and safety through advancement and development in soft robotics.

 In June 2016, a few Tata companies got together to create a first of its kind 6-year long R&D partnership with Harvard University. This is a novel attempt for Tata group to put technology and R&D into centre-stage. Soft robots for automation and as wearables are currently explored as one of the areas for the collaboration.

Light-weight, low-cost, adaptive, and safe – the soft robot has winning characteristics in its kitty. But there are still tough challenges to be overcome – efficient control, speed, and industrial robustness, among other things. But with the leaps and bounds that soft robotic research has taken in the last 3 years, I am confident that these challenges will be adequately met, and the soft robot will bring a tipping point for human-robot collaboration.

Now, re-imagine the industrial robot.