A new industry report from the Sikich Industry Pulse: Manufacturing and Distribution survey found that when it comes to automating, companies are focused on the factory floor. When asked which department in their company is currently implementing the most automation, 40% of respondents cited the factory floor. Indeed, factory automation has a long legacy of implementation in multiple manufacturing environments. The introduction of robotics to automate repetitive tasks requiring precision and error-free repetition has improved product quality immensely, driven down costs, and enabled improved and more accurate data capture.
I suspect the other 60% of respondents may have cited automation beyond the factory floor, with such process-heavy areas as customer service, sourcing, inventory management, lead generation, and a host of other areas requiring human input and interaction as a major focus. As Debbie Altham explained to Aerospace Manufacturing & Design in its article, “Automation is More Than Robots on the Factory Floor,” automating processes in these back-office and customer-facing spaces are no less complex or impactful than repetitive tasks such as placing a weld on a chassis, inserting miniscule components onto a circuit board, or precisely applying a solution to a film.
A Collaborative Process
Although the current wave of so-called generative AI (think ChatGPT) is prophesied to transform industries and the future of work, process automaton at its core must be a collaborative effort between human and machine. No process can effectively be automated unless it is first defined and optimized by the human, and then usually refined after multiple iterations.
Understand and Simplify Before You Automate
My most recent association was with a company at which process simplification was imperative in every aspect of its business. Ingrained in every associate was a methodology with the acronym “USa,” in which everyone was trained. “U” stands for “Understand.” Before any action can be taken regarding process transformation, one needs to understand thoroughly the process to be addressed. This means inserting oneself into the current process completely, to “be the thing” and follow every step through from beginning to end. “S” stands for “Simplify,” to either modify or eliminate unnecessary steps or, if needed, reimagine the process completely. The “a” stands for “Automate.” In the USa acronym, “a” is deliberately lower case to emphasize that the automation step is secondary and subordinate, and therefore must follow understanding and simplification to be effective.
A friend of mine is a store manager for a large grocery chain. He told me that an automated inventory replenishment system was recently implemented across all stores that automatically placed orders to replenish stock items, relieving department managers from having to perform this task. His anecdote included bananas as an example, as bananas happen to be the single highest volume item moving through a grocery store (who would have thought?). After several weeks of stock-outs or overstock of bananas, store managers were told to ignore the system’s ordering recommendations until the algorithms could be updated with revised human input. This is an excellent example of the need to understand a process, its inputs, and its environment thoroughly before attempting to automate it. It’s also indicative of the iterative process required to get it right.
A while back, I was one of the executive sponsors of an ERP implementation. During the “understand” phase of the project, one of the prospective users of the new system expressed the importance of automating the issuance of credits to customers which at the time were occurring in very large volume. As the team dug further into the existing process, we came to understand that complex and manually implemented pricing rules generated errors that had to be corrected via credits back to customers. We were subsequently able to simplify the pricing process and automate the remaining steps using pricing tables and algorithms that the machine could use to calculate correct pricing when orders were placed. Collaboration between human and machine allowed the machine to do the heavy lifting and minimize error.
The Human Impact
Much has been said and written about the business, employment, and societal impacts of automation, and not all of it positive. Regardless of the negative hype and fear, machines, robots, computers, and algorithms are merely tools created by humans and dependent on humans. Successful deployment of automated solutions can only be driven by knowledgeable human input and partnership between human and machine. And a knowledgeable and educated workforce capable of the analysis needed to feed the machine will never become obsolete.
Humans must necessarily adapt. As Peter Drucker said, “If there is one thing certain under automation, it is that the job will change radically and often. ”
