A few years ago, I met a capital projects and product line supervisor who was frustrated with his contract manufacturer’s performance, and even more frustrated that he and his team couldn’t see issues until it was “already too late to do anything about it.” Each day bought unforeseen challenges – a machine with unplanned downtime causing a bottleneck, a new input material that necessitated a process change, disruptions to staffing levels slowing down production, and so much more. Until one day, I showed him how a software only solution, installed at the equipment edge of his contract manufacturing facilities could not only provide real-time visibility to these challenges as they occurred, but also predict and take action to eliminate issues before they occur. Now, he can take advantage of a flexible and easy to use software platform to build and scale edge-first, autonomous applications against the data he was already generating. Now, he can see issues that have been corrected and augment the know-how of his subject matter experts with the super-human power of a platform that revolutionizes human-machine interaction. By getting started with a future-ready, autonomous-ready platform, he and his team are well on their way towards realizing the “factory of the future” that others only dream about. By taking incremental steps along their industrial digital transformation journey, with an autonomous-ready platform, he is seeing exponential gains that years ago seemed out of reach.
This is just one example. I speak with manufacturing executives and CEOs often about the good, bad, and ugly issues involved in their industrial digital transformations, and I’ve found a few common threads that separate leaders from laggards. In this post, I’ll share how we define Autonomous Manufacturing, provide some parallels to self-driving technology, and share some tips on how to evaluate and select offerings that can accelerate your journey.
Defining Autonomous Manufacturing
Industrial digital transformation leaders are chasing an elusive, and often misconstrued, vision of Autonomous Manufacturing – where humans and machines seamlessly interact to maximize efficiency, where AI can see around corners, where supply chains and equipment can adapt automatically to changes on the fly. Therefore, I think it’s important to first define what is meant by “Autonomous.” Firstly, Autonomous and Automation are related but different. Automation is using processes and machinery – often computer software and robotic equipment – to achieve levels of output that outperform human capacity alone. Automation is the first step on the journey to autonomous systems adapt to unanticipated changes. I prefer how Watson & Scheidt (2005) define autonomy: “Systems that – without manual (human) intervention – can change their behaviour in response to unanticipated events during operation.” Throw a curveball, and the system responds.
The Industrial Transformation Journey
The second concept worth exploring is that of the industrial digital transformation “journey” itself. It truly is a journey – with phases that take time, with fits and starts, homeruns, and strikeouts. Though what separates leaders from laggards is an understanding of the interconnected ways that people, processes, and systems accelerate (or slow down) each phase of the journey. This interconnectedness is particularly important when considering the “systems of record” that you invest in along the way – and leaders opt for future-ready systems that can help them in later phases, knowing that you can never get to fully autonomous operations unless your systems are built for autonomy from day 1.
Like the Journey to Autonomous Self-Driving Cars
I like to use the analogy of self-driving cars to paint the picture of the interconnected people, processes, and systems that will enable autonomous manufacturing. To achieve level 5 self-driving autonomy – where no human operator is needed – requires a journey to build, prove, and scale a number of critical components: from a real-time sensor ingestion system, to AI-powered models (built by human subject matter experts) that can “see” and react to the world around them, to a data management infrastructure that bridges in-car and cloud-based capabilities, to failsafe mechanisms for unknown scenarios. And there are myriad other pieces of the total solution puzzle including navigating a regulatory mechanism for liability and compliance. In the same way, the Autonomous Factory will require systems designed for automation, hybrid-cloud data management, advanced AI techniques and algorithms, and the right mix of people, processes, and systems that can harmonize the best of human know-how with machine can-do.
Designed for Autonomous Operations from Day-1
One common pitfall I see is over-investment in point solutions – many that claim to leverage advanced AI – that can’t enable truly autonomous manufacturing. The reason these solutions fall short is that they don’t work at operational feedback speed, and don’t integrate easily with the control systems necessary to act on their AI-powered predictions. Therefore, industrial digital transformation leaders need to consider solutions designed for autonomous operations – and this is what makes Adapdix EdgeOps so unique. EdgeOps is designed from the ground-up to enable autonomous manufacturing, though built to help leaders along each stage of their transformation journey. While we have advanced, AI-powered capabilities that customers can take advantage of today, just as important is how we help you get started from where you are. For many, this begins with real-time data ingestion, time-synchronization, and stitching across your production environment. Leveraging our DataMesh technology we can quickly move from visibility to predictive insights, and then begin to fully automate certain processes and sub-processes that form the building blocks of your autonomous manufacturing journey.
If you’re feeling confused or stuck in your industrial transformation journey, let’s chat about Adapdix EdgeOps.
Anthony Hill, Founder & CEO, Adapdix
