Autonomation

Defining ‘Autonomation’

Autonomation means “intelligent automation” or “humanized automation”. In practice, it means that an automated process is sufficiently “aware” of itself so that it will:

Detect process malfunctions or product defects
Stop itself
Alert the operator

A future goal of autonomation is self-correction. This is generally not justified by current cost-benefit analyses.

Examples

Autonomation deals with product defects and process malfunctions.

Product Defect: A press shapes a piece of metal. Sometimes the piece breaks under the force. Manual inspection would require a worker to look at each piece to see if it broke.

Autonomation would move the output piece onto a jig with a weigh scale. Then, if the piece does not fit securely, or if the weight is out of tolerance, the press would stop and an alarm would alert the worker.

Process Malfunction: One machine places output pieces onto a conveyor belt. The pieces move into a second machine – the press from the first example – for further work. Sometimes the press stops. If the first machine and the conveyor belt continue running, the pieces will pile up and spill somewhere. Manual inspection would require a worker to watch the press, and then stop both the first machine and the conveyor belt.

Autonomation would receive a signal from the press, or notice (by weight, perhaps) that the previous piece was not picked up by the press. Both the conveyor belt and the first machine would stop, and an alarm would alert the worker.

Problems Without Automated Error Detection

It is difficult for workers to detect all defects, even when alert. Some of this depends on the machine cycle times, and on the type of testing required.

It is difficult for workers to remain alert and engaged in detecting defects for a full shift. Workers generally are more satisfied in productive tasks than in full-time error-detection roles.

Depending on the cycle time of a machine or process, the worker may not be able to react quickly enough to prevent damage.

Continuing to run a flawed process will send defective work-in-progress down the line. More cost will be incurred processing these products which ultimately will be scrapped or re-worked.

Continuing to run a flawed process may damage more equipment – pieces may be jammed into a receiving machine, for example.

Continuing to run a flawed process may damage work-in-progress, if pieces fall off a conveyor belt because the receiving machine is jammed.

If the worker’s primary focus is on meeting a production quota, there will be little incentive to detect defects and correct the problems. Even if quality control and error detection are rewarded, the simple act of working to produce something will take some of the worker’s attention away from error detection.

Other Improvements With Automated Error Detection

As more manufacturing and error detecting tasks become automated, it becomes possible for workers to “supervise” multiple machines. This usually involves setting a worker in a cluster of machines, so that the worker has easy access to each machine.

Summary

Autonomation:

Improves the speed of detecting defects
Reduces costs by reducing damage to work-in-progress and equipment, and by preventing further processing on flawed work-in-progress
Improves operator morale, particularly if the operator is trained to resolve problems (rather than simply calling for a technician)
May reduce direct labor costs by permitting one worker to “supervise” several machines

Oskar Olofsson, 2009

Pages with the most "likes":

1 5S PowerPoint - Five Steps to a Better Workplace

2 Kaizen - The zen of doing it better, and making it better

3 Z-value: Setting the Standard

4 Kanban Calculator- A Card to Pull Production

5 MTBF and MTBR

I am a Swedish-based Lean consultant, and the owner of the World-Class-Manufacturing.com web site.

Contact Oskar Olofsson

Contact