4 Comments

Robots have been a promise and a fear for the last century. Up till now robots have been used for what have been called the 3 D’s — Dull, Dirty and Dangerous work. Things like building cars, vacuuming, mining, chopping up carcasses, search & rescue, and joining the armed forces. But there are also robotic footballers, pool sharks, penguins, spiders (yay!) and kung fu fighters. Robots are also starting to drive around town or do experiments (thankfully not yet on us). There may be over 8 million robots already out there.

We haven’t yet got to Rosie the Robot Maid, positronic brains, or Skynet, but we seem to be heading towards at least some of them. Robotic office workers are on the way – oh, here was I thinking some of them had infiltrated our work spaces long ago.

In this blog posting I’m introducing Aridane’s webshot; an overview of the drivers, trends, challenges and opportunities covered by the post:

  • Drivers (influencers of change): demographics (aging populations), technology, defence (warfare), economics (labour costs & productivity)
  • Trends: robots being used for more types of applications and in more complex situations; moving from programmed machines to learning machines
  • Challenges: creating versatile and fully autonomous safe robots, what types of robots will we accept safety, making robots affordable for the middle classes
  • Opportunities: improved productivity and safety, filling workforce gaps

 This post (Part 1) covers the trends, Part 2 looks at some of the challenges.

Current situation

The last few years have seen steady development and deployment or robots in a variety of settings. Industrial robots are making a comeback after two years of slow growth following the global financial crisis. The latest forecast from World Robotics concludes that there will be 1.3 million robots working away in factories by 2014. Current numbers are just over 1 million. Most of the industrial robots are employed in the electronics and automotive sectors, with Asia being where most of them reside. The US is concerned that it is falling behind [PDF, 1.32 MB] in the field of industrial robots.

Robots used in other settings — what World Robotics call ‘service’ robots — are also increasing. These are predominantly military (6,000 sold last year) and surprisingly (to me) milking applications (just over 4,000 sold). The latter are big in Europe, with New Zealand and Australia just starting to get interested in similar machines. Future farmers may need advanced degrees in engineering. More on military robots later.

Service robots for medical and logistical (such as moving freight around) applications each sold about 1,000 units in 2010. The World Robotics report notes only small sales of cleaning bots, but iRobot states that more than 6 million home cleaning robots have been sold. However, World Robotics predicts that over 14 million service robots will be sold over the next 4-5 years.

 

Drivers for Robotics

Key drivers (or influencers of change) in the field of robotics are ageing populations (leading to fewer human workers and more older people to look after), rising labour costs, reducing combat losses and increasing combat effectiveness, and technological developments (in mechatronics, materials science, sensing technologies, and of course faster cheaper computing power).

 

Trends

World Robotics notes that there are a growing number of more versatile industrial robots — ones able to do several tasks rather than just one. This is being driven by the increasing flexibility of some manufacturing processes, where product lines change quickly and products are personalised for individual customers.

So when will we get a versatile domestic robot, be able to call up ‘Ms Green Robot’ to work in our garden, or get ‘Hire a Botty’ to come over and do some home maintenance?

Not for some time. But just getting robots away from a factory floor or lab is a big achievement. And home and medical care robots don’t seem too far away.

We shouldn’t just be imagining a single house robot either. Robots swarms are likely to become more common in the near future. These can help map environments, as well as undertake surveillance. And with more things getting connected to the internet (see my previous post on the Internet of Things) your future Roomba may be conversing with other small autonomous devices to organise cleaning the home and other tedious tasks.

While the increasing dexterousness of robots and their sensory capabilities are impressive, the more significant trend in robotics is the changing approach to programming. Software is moving away from coding simple stimulus-response actions to more evolutionary behaviours, enabling robots to learn in new environments. Advances in artificial intelligence are likely to result in more sophisticated robot behaviours in the next few years. The August edition of National Geographic provides an overview of some recent advances in making robots more sociable.

There is also a move (as in other areas of ICT) for robotics to adopt an open source software approach to stimulate developments and new applications.

The larger underlying trend is of increasing automation in our lives. Dishwashers, microwave ovens, smart phones, robot vacuum cleaners, robotic genome sequencers, and computer assisted driving (and flying) are all now common. They free us from some activities or help us do others better. In the short to medium term it seems that robots will simply continue this trend. Some consider that a popular future robot could be a self propelling equivalent of a smart phone or tablet computer that acts like a personal assistant.

There will be further blurring of the boundaries between humans and machines as bionic prosthetics become more common. See this video on TED about human exoskeletons.

There are of course fears that robots will take the jobs of humans, but will this be any different from other technologies (think of ATMs, shipping containers, the internet). New types of jobs for humanoids usually emerge.

A transformation will occur when robots start making complex decisions and actions that can’t be pre-programmed. Then perhaps we’ll stop considering them as just devices and start thinking of them as creatures. (Roomba owners can get very attached to them, but it’s not quite the same thing). The goal of the RoboCup is to have a robot team beat humans by 2050. That will be a significant turning point.

Part 2 gives a brief overview of challenges facing robotics, and more on military robots.