Prior to the work of W. Edwards Deming[1], quality was generally achieved through inspection. At different points in a manufacturing line, inspectors would look for parts that deviated from what was expected. Deming adapted manufacturing to instead focus on process control, which led to today’s methods and standards for ensuring quality.
While people are any company’s greatest asset, they are not infallible. They do their very best every day, but they still make mistakes. While those mistakes are often inconsequential, sometimes they are not. Despite industry’s focus on process control, when people are involved directly in implementing a process, even guided with clearly defined procedural steps, they sometimes just don’t get it right.
People are smart. They learn through practice, and they tend to get used to the “mechanical feel” of a procedure, having gone through the same steps many times. They simply cannot help largely committing the steps to memory. Eventually, the “mechanical feel” makes it quite natural for them to put less reliance on following the procedure literally, one step at a time.
Just because people are smart doesn’t mean they don’t make mistakes when performing procedures. When they do, those mistakes fall into several categories:
Omission. Whether committed to memory or not, it is inevitable that distraction, pressure, stress, or simple oversight might cause people to sometimes forget a step. Even a strong “mechanical feel” for a procedure doesn’t prevent omission of a step. It is an inadvertent mistake that introduces potential for the procedure to fail or be compromised.
Out of Order. Failing to follow steps in order is very common, especially when a procedure is oft performed. The “mechanical feel” for the procedure can betray the person. We see this every day as we type … typos are most often letters inadvertently entered in the wrong order, even though we’ve typed that same word a million times before. The same is true for procedure steps, especially if committed to memory—another inadvertent mistake that introduces potential for the procedure to fail or be compromised.
Simple Error. Virtually any step in a procedure can be done incorrectly. Correction of such an error, if attempted, might yield unpredictable results, and cause the procedure to fail or be compromised.
Procedure Change. Once people have developed that “mechanical feel” for a procedure, and committed large portions to memory, they become inadvertently resistant to change. If an organization changes a procedure, as does happen from time to time, that “mechanical feel” has people pre-programmed to continue the old way – not out of malice, but out of habit. In fact, the less people think about the procedure, the more likely they are to do it the old way. Thus, when a procedure changes, it is critical that everyone involved try to be fully engaged in following the new proscribed steps … until a new “mechanical feel” replaces the old.
Bonus Steps. People are a clever bunch. Whether deliberate or not, when they follow procedures, it is not uncommon that they insert a step that is not part of the procedure. This improvisation may stem from distraction, daydreaming, or even intuition. But, obviously, it is not the worker’s responsibility to deviate from the procedure.
Failure to Finish. If you’ve ever built a piece of assemble-it-yourself furniture, it is very tempting to stop following the directions when “it looks like a bookshelf”. For some reason, there is a great human temptation to stop a procedure before the final steps have been completed. In a manufacturing environment, that’s bad because it threatens the consistency of product being produced.
Manufacturing processes are usually well-scripted for an obvious reason—the manufacturer wants every product to turn out the same. The procedures are commonly step-oriented, and often require the assembler to tick a box or apply initials when a step is done. It doesn’t mean that the human assembler won’t do three steps and then tick three boxes, or tick three boxes and then do three steps. But, if the procedure is properly followed and documented, the result should be a consistent product with completed manufacturing documentation.
Service is a bit different. Presuming all the products were manufactured the same way, and assuming that no product is perfect, at some point there will be failures. Traditionally, service has been reactive to those failures. An inflamed customer calls in and complains about something not working; field service is dispatched to diagnose the problem, perhaps using a troubleshooting guide from the service manual or using his/her own prior experience; and then, if possible, field service repairs the problem. Sometimes a part is needed that the service engineer is not carrying, so repair might be delayed, and a second visit required to complete the work.
Unfortunately, service procedures are usually performed under high pressure. First, the customer has an issue, and the service engineer is the one person who can make that issue go away—the customer is often breathing down their neck. Second, each service engineer has other customers with issues that they are not currently working on while fixing the present problem. That puts a lot of pressure on engineers to solve this problem so they can move on to the next.
Such pressures mean that it is common for service procedures to be incompletely followed, with steps skipped or reordered. Memory becomes important, folklore steps in, and “what I did last time” may trump factory-authorized procedures as “best approach for repair in this case”. The “mechanical feel” of prior experience is very strong. This all compounds the risk of mistakes being made or shortcuts being taken.
Manufacturing. The mitigations for human error in manufacturing are usually preventative in nature. The processes must be changed to limit the impact of people on the quality of the product produced. Automation is a common approach, with robotic assembly as the most visible method for automating assembly. As an example, it is undeniable that robotics has clearly changed car manufacturing forever, and increased product quality at the same time.
Service. Automation is also key in mitigating human error in service. For any product that is computer-driven, having the manufacturer’s technical support team perform as much of the diagnosis and repair as possible, using workflows and scripts that are automated, is clearly advantageous. It ensures a level of process reproducibility and quality that is just not achievable with people in the loop. Service automation has the advantage of increasing the First Time Repair Rate and Customer Satisfaction. It saves field engineer and service technician time, freeing them to do other tasks. As a side benefit, because service automation reduces travel time, parts waste, and parts shipping costs, it also directly reduces carbon footprint.
Mistakes are real, and they happen every day. Our best chance to mitigate those mistakes in manufacturing and service is to lessen their impact through automation. Every process that is automated reduces the impact of people on those processes, the desired benefit.
Of course, because people design processes and create automations, they too are prone to human error. But with careful process validation, those issues should be shaken out early on, allowing the process to then be repeated without further error.
maiLink™ SRM software is a service relationship management platform that automates repetitive service tasks and helps you build a rich database about your installed devices. It seamlessly integrates telemetry from your products into a database rich with profound product insights. maiLink has no per-user fee—any employee you authorize can have access to the data. To learn more about maiLink software, visit www.maiData.io or email info@maidata.io.
[1] W Edwards Deming, Wikipedia.