Article #26 - Operational Data – An Untapped Goldmine
It is becoming more apparent that companies are treating their operational data with the “respect” deserved. About time and if not, then they should start now. Operational Data is typically stored in the CMMS (Computer Maintenance Management System) or historical records repositories.
Data is knowledge. Data is power. Data is the basis of good decision making.
This article also suggests that data could be revenue generator. Therefore, a company’s data is akin to extracting gold from a mine. Like gold, data requires hard work to get and hold on to. Yet, it is highly lucrative when exploited, refined, and traded. The organization running the equipment generates the Operational Data. It can also be generated by a third party. In the following paragraphs, we see the untapped opportunities data and historical records can offer.
Third party data collection
In some industries, third parties manage and conduct maintenance. An operator may opt to save money and resources by not investing in a maintenance workforce. All or part of the maintenance work would therefore be contracted out to an external service provider. This might be cheaper, yet it could be a great loss to the operator. Especially if there is no effort to recover the records generated over time. In essence, all the learnings from the data goes to the third party. What if the contractor’s maintenance practices could be improved? Or different value-added decisions potentially made with this data? Or if the service provider is not using the data to its full potential? Therefore, companies should have full control of their data. And if subcontracting their maintenance services, have analysts overseeing the recovery and use of the data. And obviously retain all the records generated in their own repositories.
Collaborating with equipment manufacturers
Generally speaking, manufacturers design and test equipment before they sell them to operators. They excel at this and provide us with great products. Operators, on the other hand, operate the equipment until their end of life. Each of those actors “control” a specific life interval of the equipment. Operators don’t have design records and designers don’t have operational records. Operators don’t really need design records except in some specific cases. The opportunity here is for the vendors to make use of the operator’s records. What if the vendor could access those operational records? They could work collaboratively with an operator achieving the following:
Invitation for further thoughts
This article highlights some of the opportunities missed by operators and manufacturers. Some companies chose to sell their records for profit. The underlying thought is that equipment historical records are a valuable asset. They could hold much more value than we think. There could be more opportunities than those listed above. The reader is welcome to suggest them for the benefit of the entire community.
In summary, companies should treat their data for its potential value. On the flip side, it is also important to make data entry easy and convenient the workers entering this data. And have good processes around data management . Otherwise less records or even no records will be entered in the first place.
Another great comment by Nikalus Sharpe which was posted on the Accendo Reliability website. Thanks Nik! "Thanks for the article Andre! Having spent some time working for the OEM in the mining and construction industry, we would often have that great relationship with our customers where they would share their operational data and improvements would be made at both levels. There were some who were reluctant at first as they thought it could be “used against them” in someway but once we sat down and talked them through how it is of benefit, most turned around. Some additional benefits that weren’t mentioned above:
continued .... 3. Lab126 fudged Weibull warranty returns estimates because Weibull didn’t account for WEAP (Warranty Expiration Anticipation Phenomenon). 4. Recent AccendoReliability article said electronic devices are so reliable that we hardly need reliability because most survive. Schweitzer Engineering Labs EEPROMS seemed to be losing their little memories in the new year. Reliability “Broom” charts showed EEPROM reliability was fine until the new year. EEPROM vendor has shrunk the die size. 5. FAA requires lifetime data for implantable medical devices by name and serial number. I told FAA, that ships and returns counts were statistically sufficient to make nonparametric reliability estimates and protect patient privacy. Medtronic used to publish pacemaker reliability functions and periodic ships and failure counts. I used ships and failure counts to estimate pacemaker reliability and verified their estimates. So Medtronic quit publishing reliability estimates and data. Wouldn’t you like to know the reliability of the products you’re considering buying or implanting?"
Comments and very informative contribution from my esteemed and resourceful colleague Larry George. Thanks Larry! This was posted on the Accendo Reliability website. "André-Michel wrote, “There could be more opportunities than those listed above.” I (Larry) agree. Feed-back nonparametric field reliability estimates to design, process, and field engineering: 1. Estimate warranty reserves and guide recalls or warranty extensions. I offered to help Mike the Apple Service accountant estimate warranty reserves. He declined. I asked how he estimated warrant reserves. He replied, “We extrapolate from last year.” I asked what if extrapolation was wrong? He replied, “We take a charge or credit depending.” 2. Sun Computer designer said, “Why do we need you? We design reliability into the products.” He didn’t know how processes, installations, environments customers treated his products. 3. Sun Computer service manager shared field ships and returns counts, and I made nonparametric reliability and failure rate function estimates for use in diagnostics and spares recommendations. Service manager said thanks, and now we’ll have to kill you.