Issue link: https://www.e-digitaleditions.com/i/1483907
www.biopharminternational.com Manufacturing and Facilities 2022 eBook BioPharm International ® 15 OperatiOns leverage machine learning data. The software com- pany had a client who hired a team of data scien- tists to solve a complex problem. This problem con- cluded with a plant-wide shutdown and halting of production. The plant employed a team of data sci- entists to find the root cause of the issue and provide a way that the plant could avoid the shutdown event. But after months of analysis, all the data scientists engaged by the plant could repor t was that, if the speed of t he conveyor increased, t hen its overa l l production increased. This of fered no resolution for solving the plant shutdowns. Lacking the expe- rience and understanding of the plant's operations and systems hindered the data scientists from being able to deduce the problem. Mach i ne lea r n i ng helps prov ide i nsight i nto a plant's operations; but without experienced person- nel, the data cannot provide solutions for improve- ment. The same plant provided its OEM's SME with the same data and received a viable solution to the problem. The SME found system operational data and KPIs that indicated an impending shutdown. This information was then used to adjust plant op- erations momentarily, pushing through the failure event to avoid a plant shutdown. The plant could then resume its normal production schedule. By outsourcing the k nowledge management of equipment, a company can begin to leverage t he data and expertise of the OEM's SMEs with a mul- tit ude of operationa l and f inancia l benef its. Col- laboration with the OEM guarantees that process k nowled ge a nd e x per t i se w i l l a lw ay s be av a i l- able. End-users can suffer when one of a select few process ex per t s leaves t hei r posit ion. However, OEMs have severa l layers of process exper ts who can step into the SME role. The knowledge within t he OEM is instit utiona lized by t he development of sta nda rd operat i ng procedures, sta nda rd ized engineering designs, training programs, and doc- umentation. Project managers, design engineers, t e c h n ic a l s u p p or t e n g i n e e r s , a n d o t h e r s w h o have a ha nd i n t he desig n, st a r t up, a nd suppor t provide the foundation of process expertise within the OEM. As such, the process knowledge and ex- pertise are not so easily diminished. Technical benefits of data analytics A major benefit of a company with long-time em- ployees is its knowledge of the equipment, includ- ing the issues and resolutions experienced during the tenure of its employees. This builds familiar- it y with the equipment and a deep understanding of the operational histor y and issue management. This t y pe of k nowledge is built into the machine learning sof t ware, allowing access to operational h istor y a nd equipment issues for a l l employees, rega rd less of t hei r exper ience or longev it y. As a resu lt, doc u ment at ion of issues a nd resolut ions ca n be accessed at a ny t i me, bui ld i ng t he opera- tional knowledge of the operator without the need for a long tenure at the company. This f unction is displayed in Figure 2, which presents an example of issue management software. Figure 2 shows an example of an issue manage- ment list maintained for a generation, storage, dis- tribution, and ozonation system used to produce and maintain US Pharmacopeia purified water. It allows for ever y issue to be documented in detail. This helps to centralize each problem the site has ex- perienced and provides another facet of knowledge FIGURE 2. List of issue history with the resolution, description, status, and monetary impact.