The print and packaging industry has been using IA (intelligent automation) in the form of pipeline automation for decades. In fact, I remember installing one of my first systems in the late 90s, based on connected hot folders. Since then, pipeline workflows like those you see in Enfocus Switch, Callas pdfToolbox, HP Site Flow, Hybrid CLOUDFLOW, Esko Automation Engine, Fiery JobFlow, Fujifilm XMF, and others have continued to expand their offering to address internal plant operations as well as connectivity with other software and hardware systems. In fact, the vast majority of automated print and packaging production workflows are successfully using pipeline workflow solutions.

Today, there are pockets of very successful AI solutions. What are the real benefits today for print and packaging production and supply chains? AI and data analytics are being applied to demand forecasting, dynamic pricing, maintenance (predictive and prescriptive), and process control, especially in flexo and high?throughput environments where small efficiency gains convert into large annual savings. Cloud?enabled production management and MIS/ERP integration provide real?time visibility of job status, capacity, and profitability, which is becoming a priority as operations grow more complex (more SKUs, smaller lots, more regulations). With many of these early successes, we are finding that today AI is still not necessarily the answer to lights-out print and packaging plant manufacturing, although as AI and “we” become smarter, the amount of available solutions will increase.

AI is constantly trending in the news. Much of that is in the form of textual and image content generation, and more recently with agentic AI bringing autonomous systems which can set goals, create actions, and complete tasks without human intervention. Building your own agentic solutions to address specific issues in business or even life can be empowering, but according to Dwight Kelly, the founder of Apago and one of the global experts on automating print and packaging workflows, “keeping up with state of the art in AI LLMs is just getting ridiculous. Today, every two weeks, something new comes out, but there are ways that I have found to work around it.”