Boards keep asking the wrong question.<\/p>\n\n\n\n
Not: \u201cDo we have an AI strategy?\u201d<\/strong> That distinction matters because what many manufacturers currently label an \u201cAI strategy\u201d is usually something far safer and far more superficial: a vendor\u2019s roadmap, a growing pile of disconnected pilots, a corporate automation program tucked into HR, Finance, or Legal, or a procurement contest to secure the latest large language model. None of these is a strategy. They are activities. And boards do not fund activity. Boards fund measurable returns.<\/p>\n\n\n\n This is the reality that rarely makes it past PowerPoint: AI does not fail in manufacturing because the models are immature. It fails because intelligence is being layered on top of systems, processes, and governance models that were never designed to learn, scale, or move at industrial speed. When data is fragmented, ownership is ambiguous, and operational decisions cannot be traceably governed, AI simply amplifies existing dysfunction.<\/p>\n\n\n\n And the destination, across industries and regions, is almost always the same:<\/p>\n\n\n\n The Wall of Disappointment<\/strong>: impressive demos, weak industrialisation, and no repeatable operational lift.<\/p>\n<\/blockquote>\n\n\n\n This post is a board-level provocation\u2014and a decision tree to fund AI profitably<\/strong> or kill it early.<\/p>\n\n\n\n AI deployments are too often pushed into HR, Finance, or Legal\u2014not because that\u2019s where the value is, but because it\u2019s where friction is lowest. The factory is harder: legacy systems, safety constraints, OT integration, and real operational risk. So it gets avoided.<\/p>\n\n\n\n The outcome is predictable and corrosive:<\/p>\n\n\n\n AI didn\u2019t fail here\u2014the deployment strategy did.<\/p>\n\n\n\n Boards are feeling competitive pressure and are right to demand results. Executives translate that pressure into urgency. But urgency without clarity is not transformation\u2014it\u2019s expensive chaos.<\/p>\n\n\n\n What that looks like in practice:<\/p>\n\n\n\n This is not a strategy. It is wishful spending with a technology label attached.<\/p>\n\n\n\n If your production environment is running on brittle PLC integrations, undocumented interfaces, vendor\u2011locked systems, and years of deferred modernisation, AI cannot reliably see the process it is meant to improve. And if it can\u2019t see the process, it cannot win.<\/p>\n\n\n\n A model is not magic. It is an amplifier.<\/p>\n\n\n\n AI exposes the truth of your operating environment. It does not conceal it.<\/p>\n\n\n\n If it runs on zeros and ones, it is IT. Yet OT is still too often pushed \u201cto the business,\u201d outside of architectural standards, lifecycle management, identity models, logging, patching discipline, and clear risk ownership.<\/p>\n\n\n\n The consequences are structural:<\/p>\n\n\n\n This is not a technology problem. It is a governance decision with predictable outcomes.<\/p>\n\n\n\n If it\u2019s 0 and 1<\/strong>, it\u2019s IT.<\/p>\n<\/blockquote>\n\n\n\n AI introduces failure modes that most governance models were never built to handle: hallucinations, data leakage, model drift, untraceable decisions, and unclear liability. When controls cannot adapt, the default response becomes \u201cno.\u201d<\/p>\n\n\n\n And when \u201cno\u201d is the default, the business routes around governance. That is how localised risk becomes systemic risk\u2014precisely the opposite of what boards expect from control functions.<\/p>\n\n\n\n Even when pilots succeed, they often die quietly because no one can answer a basic question in one sentence:<\/p>\n\n\n\n Who owns the model in production\u2014by name?<\/em> When ownership is ambiguous, accountability disappears. And without accountability, scaling is impossible\u2014no matter how promising the pilot looked in the demo.<\/p>\n\n\n\n The most dangerous competitors today\u2014particularly fast\u2011scaling Asian manufacturers\u2014are not winning because they buy better machines. They are winning because they were built for speed from day one. These organisations are often born digital<\/em>: processes are standardised by design, data is captured as part of the operation, automation is engineered in rather than bolted on, and modern platforms and ecosystems replace brittle, bespoke stacks. As a result, their iteration cycles are materially faster.<\/p>\n\n\n\n This changes the basis of competition. Incumbents are no longer competing purely on labour cost, scale, or footprint. They are competing on learning speed<\/strong>: how quickly they can detect what is happening, decide what to change, act safely in the operation, and improve again\u2014cycle after cycle.<\/p>\n\n\n\n AI, in that context, is not the strategy. It is a force multiplier for learning speed. But like any multiplier, it only works when there is something solid to multiply. Without disciplined processes, decision\u2011grade data, and clear operational accountability, AI doesn\u2019t accelerate advantage\u2014it accelerates exposure<\/p>\n\n\n\n Every AI initiative\u2014whether GenAI, computer vision, predictive maintenance, optimisation, or autonomous agents\u2014should pass through the same decision filter. Not a slide deck. Not a demo. A decision tree that treats AI like a capital allocation choice, not a technology fashion.<\/p>\n\n\n\n Below is the logic in plain board language.<\/p>\n\n\n\n If the initiative does not move a clearly defined operational constraint, it is not fundable. Full stop.<\/p>\n\n\n\n That constraint must show up in the metrics boards already care about: OEE, downtime, yield, scrap, energy intensity, safety incidents, warranty costs, and inventory turns. If the value cannot be expressed in those terms, it does not belong in an investment conversation.<\/p>\n\n\n\n No measurement means no funding.<\/p>\n\n\n\n Without a baseline, you cannot prove improvement. An initiative without all three is not an initiative. It is a story\u2014interesting perhaps, but not investable.<\/p>\n\n\n\n Data is not \u201cavailable\u201d because someone can extract a spreadsheet. For AI to work in production, boards should insist on decision\u2011grade visibility: accurate timestamps, operational context, traceable lineage, quality controls, and repeatable collection.<\/p>\n\n\n\n If the system cannot see the process reliably, AI funding should stop\u2014and be redirected into data foundations. This is not a delay; it is risk control.<\/p>\n\n\n\n Manufacturing is not social media. Incorrect outputs can stop production lines, generate scrap, introduce safety risks, or create regulatory exposure.<\/p>\n\n\n\n If action boundaries are not explicitly defined\u2014human\u2011in\u2011the\u2011loop controls, safety cases, fallback procedures\u2014then the system is not production\u2011ready, regardless of model performance.<\/p>\n\n\n\n Not \u201ca team.\u201d A named individual must be accountable for operational outcomes once the model is live. Someone who signs off on performance, risk, and impact.<\/p>\n\n\n\n No owner means guaranteed failure. Accountability is not optional at scale.<\/p>\n\n\n\n A pilot that cannot scale is not a success\u2014it is a prototype. And prototypes are not a strategy.<\/p>\n\n\n\n A fundable pilot must produce reusable assets: an architecture pattern, a data contract, a deployment template, and lifecycle controls (MLOps). If success cannot be repeated across plants, it does not justify continued investment.<\/p>\n\n\n\n This is where most AI programs quietly die. Real costs accumulate fast: OT integration effort, downtime risk, security and audit controls, adoption and change management, vendor lock\u2011in, and long\u2011term operating burden.<\/p>\n\n\n\n If the business case only works in a simplified model\u2014and collapses under real operating conditions\u2014the correct board decision is to stop, redesign, or exit.<\/p>\n\n\n\n Funding AI is not about optimism. It is about disciplined allocation under constraint. This decision tree forces that discipline\u2014before money, credibility, and trust are burned.<\/p>\n\n\n\n Uncorrelated pilots<\/strong> \u201cCorporate AI first\u201d as the headline strategy<\/strong> Waiting for analysts to bless it<\/strong> Stopping funding in these areas is not conservative. It is discipline. It is the difference between allocating capital to build an advantage and spending it to feel busy while others pull ahead.<\/p>\n\n\n\n Boards must stop treating data as an undifferentiated exhaust stream. Not all data is equal, and pretending it is guarantees weak AI outcomes. What boards should mandate instead is clarity: data classification (Restricted, Confidential, Internal, Public), defined data domains with explicit ownership, enforceable quality and lineage standards, and a clear Bronze\/Silver\/Gold discipline\u2014where \u201cGold\u201d means decision\u2011grade data products that the business is willing to act on.<\/p>\n\n\n\n If leadership cannot point to its \u201cGold\u201d data\u2014by domain, owner, and use\u2014then any AI strategy built on top of it is speculation, not investment.<\/p>\n\n\n\n Cloud Centers of Excellence and AI Centers of Excellence cannot operate as parallel tracks. When they do, scale fractures and accountability blurs. Boards should insist on a single operating model that covers scalable platform patterns, security guardrails, cost control, MLOps lifecycle management, audit controls, and deployment standards.<\/p>\n\n\n\n AI is not a side initiative or an innovation lab experiment. It is production engineering. And production engineering demands industrial\u2011grade platforms, not bespoke solutions.<\/p>\n\n\n\n This is where most organisations quietly fail. Funding AI without enforcing accountability alignment guarantees theatre. The model must be explicit: the business owns outcomes, IT enables scalable and secure platforms, OT is governed as IT, and Quality or Engineering owns acceptance criteria and operational control.<\/p>\n\n\n\n If these roles are negotiated differently for every initiative\u2014or left implicit\u2014AI becomes nobody\u2019s problem once it leaves the pilot stage.<\/p>\n\n\n\n Boards should resist allowing GenAI enthusiasm to crowd out higher\u2011confidence value pools. In manufacturing, the fastest and most durable returns often come from computer vision at the edge for inspection, safety, and traceability; predictive machine learning for maintenance and scrap reduction; and optimisation for scheduling, throughput, and energy intensity. GenAI copilots add value when knowledge work is genuinely constraining operations\u2014not when data discipline and process clarity are missing.<\/p>\n\n\n\n GenAI layered onto weak foundations does not create insight. It creates expensive storytelling.<\/p>\n\n\n\n AI only scales when OT environments meet the same governance standards as IT: defined connectivity patterns, coherent security architecture, identity and logging, disciplined patching, lifecycle controls, and sufficient network capability for sensor, vision, and edge workloads.<\/p>\n\n\n\n If OT remains outside governance, AI will scale outside governance too. And that is exactly where operational, safety, and compliance risks accumulate\u2014out of sight until they become board\u2011level incidents.<\/p>\n\n\n\n When AI delivers real value in manufacturing, it does so in places boards can measure, govern, and hold teams accountable\u2014not in abstract capability demos. The most consistent sources of ROI are operational, repeatable, and tied directly to P&L performance:<\/p>\n\n\n\n In all of these areas, AI earns its keep by accelerating decisions under constraint\u2014not by replacing judgment, but by making it faster and more reliable.<\/p>\n\n\n\n The advantage in 2026 is not \u201chaving AI.\u201d<\/em> It is converting AI into repeatable operational lift\u2014under real industrial, safety, and governance constraints\u2014faster than your competitors.<\/p>\n<\/blockquote>\n\n\n\n If your AI strategy is mostly about tools, pilots, and demos, you don\u2019t have an AI strategy. Disclaimer<\/strong> <\/p>\n","protected":false},"excerpt":{"rendered":" Boards keep asking the wrong question. Not: \u201cDo we have an AI strategy?\u201dBut: \u201cWhich operational constraints will AI move\u2014by how much, by […]<\/p>\n","protected":false},"author":1,"featured_media":4547,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_feature_clip_id":0,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2},"jetpack_post_was_ever_published":false},"categories":[25],"tags":[483,487,480,494,490,495,488,491,481,496,485,482,492,486,493,325,484,489],"class_list":["post-4546","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-artificial-intelicence","tag-ai-governance-manufacturing","tag-ai-roi-manufacturing","tag-ai-strategy-in-manufacturing","tag-board-governance","tag-computer-vision-manufacturing","tag-executive-decision-making","tag-factory-automation","tag-generative-ai-manufacturing","tag-industrial-ai-strategy","tag-industrial-operations","tag-industry-4-0","tag-manufacturing-ai","tag-manufacturing-data-strategy","tag-manufacturing-digital-transformation","tag-mlops-manufacturing","tag-operational-excellence","tag-ot-it-convergence","tag-predictive-maintenance"],"yoast_head":"\n
But: \u201cWhich operational constraints will AI move\u2014by how much, by when\u2014and under what accountability?\u201d<\/strong><\/p>\n\n\n\n\n
\n\n\n\nThe Forces Destroying AI ROI in Manufacturing (Right Now)<\/strong><\/h4>\n\n\n\n
1) AI gets trapped \u201cupstairs\u201d in corporate functions<\/h6>\n\n\n\n
\n
2) Business urgency is high. AI literacy is not.<\/h6>\n\n\n\n
\n
3) OT tech debt is still being ignored\u2014or worse, romanticised<\/h6>\n\n\n\n
\n
4) OT still isn\u2019t treated as IT \u2014 and that\u2019s a governance failure<\/h6>\n\n\n\n
\n
\n
5) Security and Audit are enforcing old rules in a new risk landscape<\/h6>\n\n\n\n
6) Strategy and culture are misaligned<\/h6>\n\n\n\n
\n
IT? OT? Quality? Engineering? The plant manager?<\/p>\n<\/blockquote>\n\n\n\n
\n\n\n\nThe Competitive Reality Boards Should Stop Underestimating<\/strong><\/h3>\n\n\n\n
\n\n\n\nThe Board Decision Tree: Fund AI Like a Business, Not Like a Trend<\/strong><\/h3>\n\n\n\n
1) Is it measurable?<\/h6>\n\n\n\n
2) Is there a baseline, a target delta, and a timebox?<\/h6>\n\n\n\n
Without a target delta, you cannot judge ambition.
Without a timebox, you cannot manage risk.<\/p>\n\n\n\n3) Can the system see the process reliably?<\/h6>\n\n\n\n
4) Can it act safely?<\/h6>\n\n\n\n
5) Who owns it in production\u2014by name?<\/h6>\n\n\n\n
Not \u201ca committee.\u201d<\/p>\n\n\n\n6) Can it scale across sites?<\/h6>\n\n\n\n
7) Does it pay back under real\u2011world constraints?<\/h6>\n\n\n\n
\n\n\n\nWhat Boards Must Stop Funding (Even If Everyone Is Doing It)<\/strong><\/h3>\n\n\n\n
Proofs of concept that live in isolation\u2014untethered from the target operating model, unclear on governance and ownership, disconnected from scale patterns, production architecture, or business\u2011unit economics\u2014are not experimentation. They are corporate entertainment. They consume budget, create internal noise, and deliver little more than demo fatigue. Boards should treat these as what they are: cost with no credible path to return.<\/p>\n\n\n\n
Corporate automation can reduce friction and cost at the margins. But positioned as the centrepiece of an AI strategy, it consistently fails to deliver the operational lift boards expect. Worse, it often erodes workforce trust and reinforces the narrative that AI is about efficiency cuts rather than competitive advantage. That is not a foundation for sustainable transformation.<\/p>\n\n\n\n
A strategy that begins with \u201cwe\u2019ll move once the analysts confirm\u201d is not risk\u2011managed\u2014it is risk\u2011outsourced. By definition, external validation lags reality. If a board chooses to act only after consensus forms, it has already accepted permanent catch\u2011up as its competitive posture.<\/p>\n\n\n\n
\n\n\n\nWhat Boards Should Fund First <\/strong><\/h3>\n\n\n\n
The Foundations That Actually Make AI Profitable<\/h6>\n\n\n\n
1) Treat data strategy as a capital strategy<\/h6>\n\n\n\n
2) One operating model for Cloud and AI<\/h6>\n\n\n\n
3) Business\u2013IT partnership with equal accountability<\/h6>\n\n\n\n
4) Portfolio thinking: GenAI is not the whole game<\/h6>\n\n\n\n
5) OT is IT \u2014 non\u2011negotiable<\/h6>\n\n\n\n
\n\n\n\nWhere AI Actually Makes Money in Manufacturing (Board View)<\/strong><\/h3>\n\n\n\n
\n
\n
\n\n\n\nThe Board\u2019s Non-Negotiable Questions (Use These in the Next Steering Committee)<\/strong><\/h3>\n\n\n\n
\n
\n
You have a procurement plan wearing a transformation costume.<\/p>\n<\/blockquote>\n\n\n\n
The opinions expressed herein are my own and are shared in a personal capacity. They do not reflect the views, policies, or strategic positions of my employer, its management, or any associated organisation. This article represents independent analysis informed by professional experience. AI tools (Microsoft Copilot, with cross\u2011checking using Google Gemini) were used solely as writing assistance; all conclusions, interpretations, and final content are mine.<\/p>\n\n\n\n