Recently I had a conversation with a senior technical expert at an oil company. He was proud of his company’s technical prowess and its ability to develop compelling applications to detect impending equipment breakdowns. He asserted that the company solution may not provide as early a warning or be as precise or accurate as other options, but it was good enough for the job and the company actually owned it.
I felt I couldn’t argue on that technical logic. However, the really compelling argument is not just about technical prowess. Rather, it considers the skills, experience, time and money required to develop, implement and sustain an enterprise-wide asset performance management solution!
The foregoing applies to services and products that monitor machines to predict breakdowns in manufacturing. When defining requirements for our solution, we made it clear that we wanted to supply an application that works early and accurately, is easy to use, fast to deploy, and scales to meet end customer needs. Those words are easy to write, easy to insert in a PowerPoint presentation. It is an entirely different matter to achieve them in a product.
Aligning to Existing Knowledge and Processes
I’ve always felt that applications must fit the current work processes, skills and experience of the engineers and technicians already working in manufacturing – make the tools fit the artisan (the domain experts), don’t ask the artisan to learn complex new skills. Don’t ask a reliability technician to become a data scientist. We all see the prime example Steve Jobs offered – the iPhone allowed casual users to achieve intense technical competence through a delightful user interface and a finger for a pointer/selection device.
Aspen Mtell® does not ask for intense technical prowess in chemical processing or mechanical equipment knowledge. It doesn’t call for capabilities in crafting engineering equations or using specialized data science/machine learning techniques. We expect “Joe Normal,” the guy who’s already working at the plant, to easily apply the software to his work.
If it’s simple to deploy the software on one rotating machine, then it’s simple to roll out the same solution to dozens and hundreds of assets. If what you learn on one machine is easily and completely transferred to other machines, the solution will scale rapidly, saving time and money. If rollout is fast and easy, I do not just have to apply the solution to expensive, critical machines: I can apply it to every one of the facility’s 30 boiler feedwater pumps in just a few hours. With a solution so fast and simple, I do not have to run a precise criticality analysis to determine which machines to cover, I can simply and easily blanket the whole plant applying the same safety and equipment protection to all machines – in months, not years or decades.
When evaluating an asset performance management tool, think objectively about the skills and experience your staff will need to use the solution effectively. Ask about the time and money required to extend coverage across all your critical equipment. Only then can you truly understand the real cost and scalability of the proposed solution.
Learn more in the white paper “Prescriptive Maintenance: Transforming Asset Performance Management.”