Israeli start-up teaches robots how to do their job

SmartTCP has developed a new software called SmartTCPArc that can teach manufacturing systems specific tasks easily and quickly.

Efi Lebel 88 224 (photo credit: Courtesy)
Efi Lebel 88 224
(photo credit: Courtesy)
An innovative Israeli start-up company called SmartTCP has developed a new software called SmartTCPArc that can teach manufacturing systems specific tasks easily and quickly. Manufacturing automation systems, or robots, are usually used for simple and repetitive tasks. Robots can do small-batch, complicated tasks, but the time spent teaching them what to do makes them impractical. Teaching industrial robots to perform complex tasks is a big problem; it is time consuming and needs complex set-up operation. There are very few automation solutions for complex geometries in one-off and small-batch productions. But SmartTCP has developed a technology that programs industrial robots to do complex tasks, such as arc welding, cutting, grinding and drilling, in a matter of minutes. By existing conventional methods, teaching an industrial robot to weld a 20-meter-long steel construction girder can take up to three weeks. Since the metal worker would do the welding job in two days, it is not economical to train the robot. According to SmartTCP CEO Efi Lebel, his company can train a robot in 17 minutes. "And he can then go on to do the job in just three hours, and the quality of the work is excellent," he told The Jerusalem Post. "Consequently, the SmartTCPArc will boost the use of robots in one-off and small-batch production of manufacturing work because it will effectively make automation cheaper to operate and the running in time span will be considerably shorter... [Robots] will relieve humans of the drudgery of repeated tasks, dangerous tasks such as welding work in high-rise construction projects, and tasks in toxic conditions, in say the chemical industry." This may also halt, or at least slow down, the drift of jobs from high-income countries in North America, Europe and Japan to third world countries. The global free market society is not working in favor of high-income Western countries, and the reason is cheap labor and the rapid technological advances in countries such as China and India. In the not-so-distant past, most Western economic planners did not regard cheap labor in the third world as a bad thing. They argued that the much more technologically developed West would always have an edge on the third world. As long as the manufacturing of sophisticated goods remained in the US, the UK, Canada or Germany, there was nothing to worry about, they believed. They were wrong. Technological expertise may perhaps not be drifting to South and Central America, Africa and some countries in Asia, but it is drifting toward India and China. These two countries have become technological innovators, and what is more alarming to the West, they are producing engineers at a higher rate than other Western countries. India and China have all the ingredients to produce first-class engineers and high-quality technological workers. And they are doing just that, just look at the figures. According to statistics published by the Chinese government, the number of engineering graduates in China in 2005 reached 195,000; the average annual cost of a Chinese engineers in that year was $40,000. In the US, the annual number of engineering graduates is less than 100,000 and the average annual cost of hiring an engineer is $150,000. In Israel, the comparative figures are 8,000 graduates. Hiring an engineer costs $100,000 a year. Small wonder then that there is a big drift of jobs from the "developed" to the so-called less-developed world. In the next five years, 2 million jobs are expected to move east. In contrast to past years where all were connected with the manufacture of cheap consumer goods, this batch contains many hi-tech industrial jobs and hi-tech research jobs. The new SmartTCPArc can halt this process by making automation available for such manufacturing industries that are currently dependent on intensive labor production. How does it all work? According to Lebel, the SmartTCPArc has very sophisticated algorithms, and "they can read the drawings of the finished product, and can then deduct the movements and operations that the robot will have to do to finish the job."