Fighting downtime is an endless task for companies in many industries. It hurts profit margins and company reputations, making it incredibly dangerous to successful businesses. One of the worst parts about downtime is that it can occur quite frequently, but can often go unnoticed. In a recent study, the ISA found that many companies underestimate their Total Downtime Cost by 200-300%, meaning that these companies are likely quite unaware of the damage being done to them by downtime. The only way to eliminate the cost of downtime is to first identify where and when it is occurring, making downtime tracking vital to companies who intend to improve their bottom line.
Tracking downtime can be done in many ways, some using automation, and some by hand. The methods that involve tracking downtime by hand can be quite flawed, however, as humans are often inherently less accurate than machines and may have incentives to mis-report downtime accurately.
Manual Tracking Downtime
When downtime tracking is done by employees, it is often documented by hand or on a computer in a spreadsheet. In theory, the operator documenting the downtime would write down the length of time that the system was down, the reason that it was down, which specific machine or part of the system experienced the issue, and other relevant details. However, multiple problems occur when tracking downtime this way. The most obvious problem would be that the manually transcribed data can be lost if it is not documented properly, not only when the data is written on actual paper, but even on computers files can be placed in the wrong folder or deleted accidentally. Manually entered reporting also makes it difficult for the right people to get access to the data. Data kept in spreadsheets can be difficult to gain insights from that can lead to changes and adjustments that can positively impact the business.
Many other problems come from operator input, including inaccurately logging data. Operators may simply estimate the amount of time that a system was down, or worse yet, for brief stops, the operator might not record downtime at all. If the time estimates are lower than the actual time a machine was down, or if there are short but frequent stops that get overlooked, this time can add up quickly and cost the company while remaining completely undocumented and undiagnosed.
Employees may also feel rushed to resume the work they were doing before the downtime event and may not write down enough detail about where or why the downtime occurred. If downtime is recorded but the company doesn’t know why it occurred, the problem can’t be fixed. A final and troubling inaccuracy in manual downtime tracking is the possibility of employees inaccurately logging downtime to hide excessive downtime periods. Whether this would be motivated by laziness or the stress of an employee feeling rushed to get a machine back online, it can cause companies to underestimate their downtime.
Automated Downtime Tracking
Automated downtime tracking is the answer to a lot of problems that human error can cause when tracking downtime. It takes the job out of the hands of employees, who can then focus more on operating machinery, while it also documents downtime events more accurately.
As much information as possible is needed when tracking downtime to pinpoint the exact reason that downtime is occurring, and this is very easy for an automated tracking program to do. Details about the downtime like which process area experienced downtime, which machine, which product the machine was manufacturing, what shift number was working at the time and the time at which the downtime occurred can all be quite valuable. Machine malfunctions, difficulties with manufacturing specific products, and shoddy employee work can all be highlighted as a result of these details. Better still, all of these details can be entered into an automated system in advance and tracked automatically, the product info can be entered whenever a product begins to be manufactured, the time needs to be set once and shouldn’t need to be set again, the shift number can be entered via a schedule based on the time, and the machine number and process area can be entered whenever a new station is being monitored.
One of the most important details, of course, is the reason the downtime occurred, which can sometimes be automatically collected based on an error code that the machine sends to the automated software or when accurate and timeline downtime measurements are made across interconnected machines. Other times, an operator must enter the reason the downtime occurred, which is not always perfect as it brings human error back into consideration. Just as manual downtime tracking can be flawed when operators feel the need to rush to start production again, the same can happen if they feel rushed and enter an inaccurate downtime cause into the automated software.
To prevent this from occurring, the software for entering downtime data must be as simple and easy for the employee as possible. A reasonable list of downtime reasons should be presented to an operator, not so many that common reasons are hard to find and select, but long enough so that “other” is not the top reason that downtime occurred. That would clearly not be very helpful when the company assessed the downtime data when looking for where to improve. Another good piece of data to collect would possibly be “operator comments,” in which an operator could add any additional information that may not be indicated by the other data details. If an operator is working on a machine frequently and it experiences downtime, who better to ask about what went wrong than the operator?
Once the downtime data has been collected, a Pareto chart can be made to find where the biggest causes of downtimes lie. Vilfred Pareto was an Italian economist who developed the Pareto principle, or the 80-20 rule, which states that 80% of effects come from 20% of causes. This applies to downtime because often only a small amount of problems may be responsible for the vast majority of downtime. Pareto charts take the downtime data and classify where and why downtime is taking place, to help pinpoint the issue causing the most downtime. These charts require an understanding of what types of downtime occur on each machine or in each system, and require the time recording accuracy that automated tracking brings, so that even small downtime events can be improved upon or eliminated if they are causing significant downtime issues for a company.