When I began my academic career as a lecturer in robotics and control engineering at the University of Birmingham, the BBC had just screened a Horizon film titled ‘The Robots Are Coming’ about the increasing use of robots in industry; British Leyland was still setting up Britain’s first robotic car production line at Longbridge to manufacture the Austin Mini Metro; the Science Research Council, the predecessor of the current day Engineering and Physical Sciences Research Council which supports research in engineering and the physical sciences in the UK, was about to launch its first robotics initiative; in other words, the country was being gripped by robotics fever.
However, robots were not a new phenomenon, the word ‘robot’ having been introduced in the early 1920s to the English language by the science fiction play ‘R.U.R.’ (which stands for ‘Rossum’s Universal Robots’) written by the Czech author Karel Čapek. Then, there were also robot stories by Isaac Asimov in the 1940s that were later published as the famous ‘I, Robot’ collection. The Unimate, the first industrial robot, was developed in the late 1950s. It was installed at General Motors in 1961 to handle die castings. It weighed two tons and cost US$65,000 to make, although reportedly was sold for only US$18,000.
Since that time, some 2,470,000 industrial robots have been installed in factories around the world, with between 1,235,000 and 1,500,000 units operational at the end of 2012 (Word Robotics 2013). Hardly a car production line now exists without robots performing such tasks as welding, handling and spraying. Robots have also found jobs in many other industrial sectors including the electrical/electronics, rubber and plastics and food sectors. In addition, there are millions of service robots, in professional applications (defence, healthcare, security, rescue, etc.) or personal and domestic use (homecare, vacuum cleaning, lawn mowing, entertainment, etc.).
Associated with the rise in robot numbers is the reduction in their costs which started in the 1980s. For example, back then, we paid c. £25,000 for a PUMA robot and £40,000 for a Unimate for our laboratory. Although one can no longer buy machines equivalent to the latter, for about the same price as one PUMA, we recently obtained two more powerful Kuka robots for our new Masters programme in Advanced Mechanical Engineering. We also purchased a twin-armed Baxter complete with vision sensing for even less money.
A critical factor in this robot cost reduction is the decrease in the costs and dimensions of computers and the increase in their powers. Another factor is the positive feedback that operates between cost and quantity: the more robots we produce, the lower they cost; the lower the robots cost, the more robots we use and produce, etc. This virtuous circle will prove important in the future when organisations such as Foxconn flood the market with millions more machines. The majority of them will go into jobs that have been robotised successfully, because, as I have observed over the years, once a job has been successfully robotised, it will remain so.
This leads us to a law to complement Asimov’s Three Laws of Robotics. Whereas Asimov’s laws have been violated (e.g. the First Law about not injuring a human being or allowing a human being to come to harm), given the evidence and circumstances to date - escalation in wages, rise in robot capability, fall in robot costs, need for machines to work in jobs unsafe or impossible for people – Pham’s Fourth Law of Robotics which says that “Robotisation is irreversible” is likely to be unbreakable. Thus, once robots have conquered an area, they will occupy it in perpetuity. The robots have come. They are here to stay.