Now, three years later, calls for industrial automation in all fields - including the development of robots - can be heard everywhere, and ITRI's pioneering is beginning to pay off.
The ITRI has designed and manufactured the first robot, designated as the ITRI-E model, which will make its debut at an information exhibition at the end of this year.
According to Dr. Jia-ming Shyu, director of ITRI's mechanical industry research laboratories, ITRI-E has four axes and can operate freely and accurately within a three-meter radius. It can lift an article weighing less than six kilograms.
Shyu says the ITRI-E has similar speed and response ability as the robots made in the U.S. or Japan, but notes that U.S. and Japanese robots have only three freely controllable axes. A pneumatically-controlled fourth axis is added to ITRI-E, so its performance is by no means less adept than foreign makes.
A robot made in the U.S. or Japan is priced at more than US$15,000; the ITRI-E is manufactured at a cost of US$5,500, and can be marketed at around US$7,500, giving it a strong edge on the competition.
Because robots are new in Taiwan, users still cannot free themselves from the limits of prior comprehension. Dr. Shyu says operation of a robot is so easy that only one hour of practice is required. The question is whether the user can employ the robot to his maximum advantage.
The use of robots does not necessarily call for changes in industrial equipment. A robot may replace at least two workers, and can work continuously for two or three shifts. Shyu says robots are particularly good for work in garment, food processing, plastics, electronic, textile, and hand tool plants.
The Ministry of Economic Affairs has drawn up an eight-year plan for the development of industrial robots in Taiwan. It will, through open bidding, select one or two promising manufacturers from the machinery, electronic, or electrical appliance industries, to engage in commercial production of the ITRI robot.
The manufacturers must meet the following conditions:
- Register a minimum annual taxable business revenue of NT$200 million (US$2 million) in terms of 1982 monetary value.
- Be able to send at least two mechanical and electronic engineers to ITRI for a minimum of three months to study the technology involved.
- Provide a detailed plan for the production and marketing of industrial robots.
The company or companies chosen will be eligible for incentives offered by the government to firms engaging in the development of automatic industrial tools. These include long-term, low interest loans and a tax rate of no more than 22 percent on taxable income.
In this way, the government hopes to develop 10 kinds of industrial robots for loading & unloading, welding, painting, and assembling jobs over the 1983-86 period. The plan also calls for production of four kinds of robots with sensor between 1987-90.
Technical cooperation with foreign producers, or purchase of necessary production technology from abroad, is envisaged in the plan, but transfer of such technology must be approved by ITRI in advance.
Despite this significant progress, Dr. Fang is not content with the extent of the program, placing his eye more on the development of robotics than robots. He feels that it is not overly difficult to design a robot, and that it is a much more demanding task to establish the science of robotics.
The development of robotic science involves the coincident development or various branches or science and technology, including optics, electronics, sensory mechanisms and so on. National Chiaotung University established an optical electronics department last year, a big step in Dr. Fang's development plans.
"We need at least 100 professionals to lay the foundation of our robotic development," according to Dr. Fang; cultivation of these scientists will take three to five years. "We will do the job despite the difficulties ahead," he asserts.
The development of robotics began in the United States, but owing in part to trade union resistance there, progress has been slow. The U.S. quickened its pace only after it had been overtaken by Japan. It is estimated that 3,600 robots are in use In Japan as compared with 3,500 in the United States. Last year, a Japanese robot was exhibited in Taipei, delighting visitors with its performance - serving soft drinks.
Robots can save much manpower. They can take over dangerous as well as monotonous jobs. They are man's "slaves" of the future. "British military men are sweeping the mines in the Falkland Islands at the risk or their lives. How marvelous it would be if they could assign robots to do that job," Dr. Fang commented.
The institute has now retained 10 top-notch scientists to help develop robotics here. Among them is Dr. Howard K. Lee, president of Righton Automation Systems, Inc. of the U.S., and of the Numertek Corporation, a firm based at the science industrial park in Hsinchu Numertek, though newly established, IS one of the few firms here engaging in computerized numerical controls.
Dr. Lee chose robotics as the topic of his doctoral dissertation at Purdue University in the United States.
"Americans still lead in robotic technology, but the Japanese dominate the market," Dr. Lee said. "This is because the Japanese have paid more attention to the practical use of robots in replacing human labor."
results
Dr. Lee studied the control, design and manufacture of robots under Prof. Richard Paul, a U.S. pioneer in robotics. He noted that robotics has a history of only ten years. Japan introduced its technology from the United States and has opened a new vista in robotic development. The Japanese are smart in employing robots - using them wherever they can. "But most of the robots used in Japan are simple models," noted Dr. Lee.
He divides robots into three categories. First is the intelligent model, which has sensor and vision system and can speak. The second involves machine models with computerized numerical controls, and is most popular. The third form is actually a mechanical arm, which positions and returns tools to pre-set places, and is often used by the Japanese.
In the U.S., a mechanical arm is not regarded as a robot. It is considered a mechanical tool. A mechanical arm is controlled by air pressure, and only two or three axes of movement are needed. A more complex mechanical arm may have six axes, enabling it to bend its "wrist and fingers."
U.S. robot development emphasis is on the intelligent model. But mechanical arms are needed in the U.S. market, especially for painting and welding tasks in the automotive industry. Such U.S. firms as IBM import large numbers of mechanical arms from Japan, Italy and Germany each year to supply the U.S. market. Their objective is to occupy the market first with the imported goods, and then replace the simple mechanical arms with intelligent robots.
In terms of drive system, robots can be divided into electrical, hydraulic and pneumatic models. At present, robot development in the U.S. is concentrated on the electrically-controlled model with computer vision. It can recognize an image, discerning the shape or an object. The main obstacle in U.S. development at present is the lack of coordinated progress in the mechanical and control engineering aspects. Because control is not sufficiently adept, the robot is clumsy. According to Dr. Lee, U.S. mechanical engineers could make the robots more facile by giving greater consideration to control requirements.
Dr. Lee obtained a U.S. doctorate in robotics seven years ago. He is among a very limited number or robotic scientists in the world. Last June, he returned to Taiwan to attend the National Development Seminar, and stayed on to assist in industrial automation here.
He believes that robot development in Taiwan should begin with the design and manufacture of simple mechanical arms. Considerable success can be achieved in three or four years, he says, and notes that the U.S. can be a vast market for Taiwan robots. As for development and research in robotics, efforts must be addressed to language, visual ability, controls and selection or materials.
Dr. Lee says that the Republic of China has enough electrical, mechanical, electronics and "control" engineers, but notes that they need integrative training. If they work hard, the era of robots in the Republic of China will not be far off. Japan has emphasized practical uses in its robot development programs and therefore controls the market, enabling it to supply ample capital for further development. The Republic of China should take this as an object lesson, he declares.