In 1991 researchers at the University of Cambridge in the UK hooked a camera up to their network so staff could see if the coffee pot in the breakroom had coffee or not so they could avoid making a trip when the pot was empty. In that same decade, enterprising engineers often hooked up micro-PLCs in their homes to sensors like thermocouples, limit switches, and their lights so they could monitor the home and control the lights from their computers via the internet. These early enterprising engineers bypassed the tedious programming that engineers of the ‘70s and ‘80s used with single-board computers like the SYM-1. Since that time we have come a long way. So too have modern Manufacturing Execution Systems (MES) – today there is no need to build your own MES solution.
For perspective, just look at home automation today. Your smart thermostat easily connects to your Nest or Echo as does a variety of cameras, doorbells, door locks, lights, and other devices. You control them either via your smartphone or your voice. Many of the devices have built-in AI such as my Ring doorbell that recognizes when a human is in the field instead of an animal. This brings up the question, “When did our homes become easier to automate than our factories?”
Manufacturers used to build many of the tools they needed such as stamping presses and painting systems. This made sense when they were first used in their plants – few other options existed. But today, they would never think of building most plant equipment. Yet, many of these manufacturers still are building out their information systems with IIoT platforms, separate AI and machine learning tools, scheduling packages, dashboarding applications, and even spreadsheets to help them run their operations.
For many, the argument is that their operations are “unique.” The rationale goes that off-the-shelf software won’t allow them to maintain their competitive edge – coupled with the myth that a do-it-yourself approach costs less to implement and maintain. After over 50 years in manufacturing and working in and with over 100 different manufacturers in dozens of different industries, it is my humble opinion 99% of these claims are not true.
Today’s modern MES, built on a microservices architecture, is configurable enough that it would be extremely rare that it couldn’t meet any manufacturer’s requirements. Just as my smartphone brings together mapping services, geolocation, scheduling software, and an e-commerce app to deliver a rideshare app on my phone whether I am in the US or India, today’s MES solutions can meet a manufacturer’s operational management needs in any industry, anywhere.
Invest in Manufacturing Operational Excellence, Not IT Engineering
Today there is a critical shortage of STEM skills. It just doesn’t make sense for 10,000 manufacturers to try and hire data scientists, IoT skilled programmers, user experience designers, and software engineers when the 100 or so MES vendors have invested in those same skilled people to build the same solutions you would eventually end up with. For the vendor community, hiring the best and the brightest in each critical area is a matter of competitive survival. Additionally, they are continually investing in new technology to remain competitive, rolling out upgrades almost constantly as they have moved to cloud-based SaaS deployments.
Improve the success of your MES investment by reading this article, Avoid These 3 Pitfalls for a Successful MES Install.
Manufacturers are generally better off investing their technology budgets in designing better processes and products, not in designing IT solutions. With cloud-based deployments and incremental upgrades becoming the norm for an MES solution, it would be hard to justify the IT support staff required to maintain and support a homemade MES to the same level of functionality. And, if you add in the need to integrate with new shop floor sensors, systems, and technology as well as the business systems used in the front office or the PLM systems used in product design, it should be obvious the DIY option is now not viable anymore. Yes, there will be situations where a manufacturing process is so new or different that a stopgap home built app might be needed for the short term, but any good MES provider with a microservices architecture should be able to incorporate the functionality required by the new process in short order and then support it going forward.
Just as I don’t need to write programs to control my home and monitor it from anywhere in the world, manufacturers today rarely should need to build their own MES solution to manage their manufacturing processes.
Dan Miklovic is the founder and principal analyst at Lean Manufacturing Research, LLC. He has a wealth of experience as an end-user, software vendor, consultant, and market research analyst. He led a plant applications development and implementation team at Weyerhaeuser, was a process system engineer at Scott Paper, led the network design team at a large engineering firm serving the pulp & paper and mining industries. His industry analyst experience includes roles at Gartner, Sustainable Collaborations Group, and LNS Research. He is currently a member of The Analyst Syndicate. He has authored dozens of articles, contributed to several engineering handbooks, authored a text on industrial networking, and was a co-host of World Business Review, a TV program seen on public television, CNBC, and other outlets.
You may contact Dan at [email protected].