I have consulted to a number of plants with recently acquired robots. Often the robots didn’t work very well. In some cases the robots were doing work that made little sense. In one case, the robot was taking a stamped part from a hanging line and putting it on a belt. It was so unreliable that a worker had to stand by to catch parts that were about to be dropped by the robot. Having the hanging line descend to just above the belt could have removed each part without a robot.
In a factory in Brazil, a robot cell had been installed to weld some parts together. A second cell was in the process of being installed. Management was very proud of it’s new technology. They had a five minute video of the robots in the dining room. The bright orange robotic arms waved through the air and set forth a shower of sparks with each weld. The dramatic video played endlessly, for all lunch breaks on all shifts to see.
Not a week had gone by since the robots were installed when they didn’t break down and required repair. No one in the plant could do the repair and the technicians who could do the repair were four hours away. When they arrived it often took an entire shift to get the robots up and running again.
I am not against robotic technology. It is especially valuable where the operations are dangerous for a human operator or very heavy parts must be moved into place. There are operations in which it can be justified to improve productivity. There are also a lot of cases in which we could gain a lot if we only treated people the way we treat robots.
This was driven home to me at a plant that made large metal cabinets. There were many operations in which sheets of steel, cut to size and punched with holes in the appropriate places, were manually inserted into brakes that made precision folds. The plant was installing robots on two of the brakes, as a productivity initiative. They could have gained a lot of productivity by treating the human operators who ran the brakes as they treated the robots that they were installing.
Robots require some special attention. The material on which they are to perform must be carefully placed within reach. It must be ready to go, oriented just so, or the robot will not be able to pick it up properly. There must be, also within reach, a designated place to set the finished work down, and when the space gets filled to a certain level, the finished parts must be removed, or the robot cannot continue to operate.
Much can be gained by treating the human operator in the same way. Put the material on which he is to operate within reach and optimally oriented, making it easy to pick up. Give him an easy place to set down the complete work, and remove it, then get it out of his way in a timely fashion.
Instead the brake operators in this factory had to go get their material from another operation, pushing it on heavy carts through aisles, often with the obstacles of other carts to be moved aside. When the carts with their finished work were full, they had to push them away and then hunt for an empty cart to move into position. During these activities, the brake sat idle, producing nothing. The operator was not doing any work that added value to the product when he was moving material or hunting for carts. Some time samples that we took indicated that as little as 30% of the operator’s time was actually spent on adding value.
The operator could be much more productive if there were a system by which material could be placed within reach and finished work removed on a continous basis. Depending on other non value adding factors, such as changeovers, maintenance and repairs, it is conceivable that the operator’s productivity could be tripled in a case like this.
When implementing Lean Manufacturing, we keep these three activities separate if at all possible: 1. Adding value, 2. all material handling and moving, and 3. all repair work. This does not mean that any given worker cannot do all three, but the activities are not mixed together. In any given substantial time period the worker is dedicated to one of the three.
Both repair and material handling take time away from adding value, if they are done by the production operator. The waste is hidden. Repair waste cannot be measured effectively if the repairs are sent back to the production worker to mingle repair work with production work. If repair work is done separately from production, we can see it, count it and record it. The solution is to have these functions done by workers other than those doing production, while we work at eliminateing repair work and transportation as much as possible.
A group of a dozen highly skilled workers assembled a wide variety of hydraulic cylinders, from kits of parts, using technical drawings and specifications. Each had his own work bench and tools. He would start by pulling the kit of components together in one or more plastic tubs. He would take them to a parts washer, wash all the components, and use an air hose to dry them. Then he would carry them to his work bench and begin assembly. When the assembly was done he would take the set of cylinders to a test station and put them in a queue. He would repeat the process working on another order from another customer, with a different number of cylinders of considerably different size and complexity.
The assembly department was always in backorder. At times machinists were taken off their machines to help with assembly, to campaign the backorder. This was not efficient, as the machinist did not have anywhere near the experience of the skilled assemblers, and they were taken away from their machining work.
At my urging, we conducted an experiment for a day. We took one skilled assembler out of assembly work and made him a parts supplier. He would pull together the kits, wash the parts, and get them to each assembler, putting them within easy reach, while he was working on the prior order. We told the assemblers that we would treat them like surgeons. Everything they needed would be put within reach. They should not have to go hunting for anything. Productivity jumped 25% that first day. The process was made permanent. Eventually a parts kitter and washer was hired, so that the skilled assembler could go back to using his main talents.
Non-value adding work can be concentrated in one person, or a few people. The waste is hidden and much greater if every production worker must do his own non-value adding work. Treating the workers as if they were robots, in this context, means bringing everything they need to do their work within easy reach and removing their finished work on a regular basis.
No comments:
Post a Comment