The ideal level of automation for lean systems

The ideal level of automation for lean systems
Author: Rick Harris
Published in "Lean Directions", an electronic newsletter produced by the Society of Manufacturing Engineers.

For some reason, people have absorbed the idea that lean systems require manual operations, with operators responsible for loading the machines, running the cycle, unloading the part and passing it on. However, this is a big mistake. Achieving efficient flow at the "pacemaker process" of the value stream, an indispensable element of a successful lean conversion, is practically impossible without a certain minimum level of automation.

The "trick" lies in determining the ideal level of automation. Too little automation can mean a loss of efficiency. And too much automation can mean lower reliability and less control over the system. The right level is the one at which maintenance does not consume unjustifiable resources. This can make your decisions easier and allow 95 to 100% of the uptime required by the pacemaker process to be achieved.

Level of automation

L e v e l		Machine load	Machine cycle	Machine unload	Part transfer
	1
	2		Auto
	3		Auto	Auto
	The great divide
	4	Auto	Auto	Auto
	5	Auto	Auto	Auto	Auto

Among the five possibilities relating to levels of automation, you can count on this solid reliability by positioning yourself close to what we at the Lean Enterprise Institute call Level 3 Automation (see figure). This Level provides the most efficient and flexible combination of operator and material movements. Here is why.

First, you need to understand that the "pacemaker process" must work like a soccer player who never gets injured and is never sent off in a game. In other words, it is present in every game.

The reason for this is that the "pacemaker process" is your scheduling point. It is usually where implementation of the leaner Future State, sketched out through Value Stream Mapping, begins. The "pacemaker process" is usually found at the end of the value stream, where products take their final form. Most of the time it is the final process of the entire value stream, but there are exceptions.

The process is called the "pacemaker" because the cadence at which materials flow through it is what determines the pace that the upstream operations in the value stream must follow. The accuracy with which these pace-setting operations run defines how the end customer's demand is met. Malfunctions in the "pacemaker process" compromise the entire value stream, which will never truly be lean. More people, equipment and materials will be needed to meet demand.

Exploring the Levels of Automation

With Automation Level 1, the operator is the one who loads the machine, runs the cycle, unloads the part and transfers it to the next workstation. The problems in this case are high costs and quality, which make this Level impractical. Automatic cycles are necessary to ensure minimum quality standards; moreover, manual cycles are too expensive in some regions because of high labor costs.

Level 2 already represents some improvement. Operators load the machine, whose cycle is now automatic, manually unload the finished part and transfer it, also manually, to the next process. But there are serious drawbacks. The operator will arrive at the next process carrying the part just unloaded from the first machine. The operator will have to put the part down somewhere, unload the finished part from this second machine, and load the second machine with the part still to be processed before finally running the cycle. Then the operator of the next process will have to do the same, repeating these wasted efforts.

The most obvious problem is that the operator handles more than one part at different stages of production: both the one that has not yet gone through a given process and the part ready to move on to the next operation. Sometimes managers imagine they can eliminate this waste by placing the processed parts in containers or racks. In fact, the waste is not eliminated and the problem remains unsolved. The double handling is converted into a heavier workload for two or more operators and into the inventories between processes that result from it.

Level 3, on the other hand, manages to optimize flow through the "pacemaker process" by using automation so that parts can be ejected automatically once the cycle is complete. As a result, the machine will always be empty when items coming from upstream operations reach it. Automatic unloading will eliminate approximately 50% of the handling required. Moreover, machines can be smaller, more efficient and less complex when Level 3 is used, which requires fewer resources for maintenance and technical support.

Another benefit: automatic unloading systems are not complicated to design (compared to systems where loading is also automatic), which makes Level 3 relatively inexpensive. It is the best investment strategy, because the operator loads the machine, the machine runs the cycle and ejects the part automatically. When the operator arrives with the next item, all that is needed is to load the empty position, start the machine, pick up the processed item and transfer it to the next operation.

With automation Level 4, you leave the simplicity, reliability and low costs of Level 3 behind and start dealing with greater complexity, wasted time and high expenses. The machine is loaded and unloaded automatically. The operator still manually transfers the items to the next operation. The problem with this Level is that automatic loading requires sensors or "vision systems" to position the item correctly. The operator is usually more efficient, since he can make adjustments immediately when feeding the machine.

The extra complexity that begins to appear with the use of Level 4 becomes more problematic when Level 5 is used, in which the entire process is automated. The reliability of a "pacemaker process" with this degree of automation is around 70% or less. Imagine the implications of this for the rest of the operations and, consequently, for the end customer.

Another problem with Levels of Automation beyond Level 3 is that changeover times can become too long, owing to the greater complexity involved. This leads to the need for longer shifts and larger inventories. And what is worse: if demand falls below forecast, you will be facing an economically unfavorable situation. If demand is higher than expected, your only solution will be to invest, in a hurry, in another equally automated line.

With Level 3, on the other hand, it will be possible to add operators to increase output, instead of investing in new equipment. Note, however, that if the "takt time" is only one or two seconds per part, Levels 4 and 5 should be considered real options.

It is not that automation is an evil. In reality, there is no lean system without a minimum of automation. But we are somewhat wary of over-automation, sustained by a love of technology and by our accounting systems' focus on direct labor cost - without the necessary concern for uptime and for the underutilization of installed capacity. Thus, one of the great challenges in implementing lean manufacturing is correctly determining the Level of Automation to be used. The secret is to use the right Level - Level 3.

Rick Harris
A former Toyota manufacturing manager, Rick Harris is co-author, with Mike Rother, of the workbook Creating Continuous Flow. He is also president of Harris Lean System Inc., based in Kentucky, USA).