Autonomous manufacturing: Orchestrating the agentic leap
Industries like Manufacturing, Petrochemicals and Energy are on the cusp of a transition. The era of autonomous operations – where AI suggests as well as executes actions – is emerging rapidly. However, the path forward is anything but smooth. Labor shortages in critical markets necessitate the realization of self-orchestrating production facilities capable of functioning with minimum manual intervention. Although AI is poised to deliver on this front, business leaders need to confront the impediments proactively rather than reactively. None of them is trivial, and all demand immediate attention.
Barriers to autonomous manufacturing
The state of operational data is the biggest and foremost hurdle toward achieving end-to-end autonomous operations. While new tools have accelerated data cleansing and harmonization, many factories still lack the digital foundation required for AI‑driven autonomy. According to a 2025 Microsoft-commissioned Forrester study, organizations that unify data across IT and OT systems can reduce defects as well as equipment failure frequency by up to 50%. Yet, 76% of industrial manufacturing leaders still cite unreliable data as a top AI risk according to KPMG Global’s tech report 2026, underpinning that data abundance and data quality are not the same thing.
Besides, autonomous operations cannot succeed without top‑down ownership, clear KPIs and a relentless focus on business outcomes like maintenance reliability and yield forecasting. While nearly half of the world’s industrial organizations lack confidence in their manufacturing strategy to deliver expected results over the next three years, two‑thirds are still not pursuing the aggressive operational redesign needed to support advanced automation. In other words, without clear accountability, aspirations seldom come to fruition.
Legacy infrastructure is another factor that holds manufacturers back. In this regard, the main bottlenecks are compute capacity, chip supply, unstructured data and poor governance, which directly impact the factory floor. This brings us to the cost and investment recovery. Rewiring the assets and changing the control architecture are expensive and complex endeavors. However, considering the industrial automation and platform companies (such as Cognite, ABB, AVEVA, Rockwell Automation and Palantir) are doing, businesses may reap ROI sooner than they expect.
Solving the data, governance and AI puzzle
It all starts with laying a solid data foundation, building domain‑aligned data teams with service‑level agreements and investing in data structuring. For this, aligning AI, operational and business processes must be among the top priorities.
Likewise, we must move governance beyond boardroom discussions and start operationalizing it by creating AI governance boards, implementing bias testing and establishing explainability thresholds as standard steps in their development cycles. Without these controls, scaling AI becomes a liability, especially in industries like Manufacturing, Energy and Petrochemicals.
AI as technology is evolving and gaining traction rapidly. By 2028, autonomous AI agents are projected to make 15% of decisions pertaining to everyday work. AI agents are being designed to autonomously execute complex workflows, end to end. They can be trained to reason across IT and OT environments, coordinating decisions in real time. For manufacturers, this opens the door to a step change from:
- Rule-based to learning systems
- Human-in-the-loop to human-on-the-loop
- Isolated optimization to system-wide orchestration
At its core, AI agent-powered autonomous operations sit at the intersection of four capabilities:
- Data convergence
- AI and advanced analytics
- Digital twins and simulation
- Closed-loop execution
The key is to build on a modular, flexible and secure foundation that can support these capabilities. In this regard, companies like Palantir and Cognite are already demonstrating how unified data platforms can power industrial AI. That said, no single company can solve these problems and overcome the challenges alone. OEMs, operators, owners and system integrators will all have to collaborate.
Synergizing for the ‘lights-out factory’
The future of manufacturing belongs to those who act decisively. Labor shortages will not wait. Competitive pressures will not ease. But with the right platform, the right partner and a clear focus on the fundamentals, fully autonomous operations are closer than we think.
Our AI Force platform is built specifically for the autonomous operation revolution. Being modular and flexible, it enables you to start small and scale fast. AI Force’s RAG builder and library create contextually relevant responses by connecting seamlessly to your enterprise data sources, unifying requirements, source code, test plans and tickets. It also facilitates designing and deploying AI agents tailored to specific workflows. And because AI Force works with both LLMs and SLMs, you remain in control of your AI.
At HCLTech, we are not just building technology; we are building the bridge to your autonomous future. Let us cross it together.
