New Delhi, July 10 -- India's manufacturing sector is entering a new phase where competitive advantage is shaped not just by production capacity, but by intelligence. In the steel industry-where energy costs, asset utilisation and production efficiency directly influence profitability-AI, Industrial IoT and connected-factory technologies are redefining how plants operate and how decisions are made. As manufacturers move beyond automation towards intelligent operations, data is becoming the new differentiator. In this edition of TechSarthi, Anoop Kubba, Customer Advisory Head at TechCircle, speaks with Prakash Tripathi, CDO & Senior Vice President - IT, Shyam Steel Industries, about building an AI-ready manufacturing enterprise, breaking IT-OT silos, and why trusted data-not just AI-will determine the industry's next wave of growth. Edited excerpts.
Manufacturing Enters Its Intelligence Era
Anoop Kubba: Prakash, let's jump straight into the big picture. Manufacturing has long optimised for raw mechanical efficiency. Today, everyone is talking about autonomous operations and the "AI era." From where you sit, is this marketing noise - or a permanent structural shift in how steel is made?
Prakash Tripathi: It is absolutely a fundamental shift, Anoop - but we have to look past the buzzwords. For years, our plants relied on automation to do things faster with less manual effort. What AI brings is not just speed; it is cognitive depth. We are moving toward systems that analyse thousands of operational data points, learn from fluctuations and actively assist our engineers in making complex decisions.
In steel, AI is already visible in predictive maintenance, quality inspection and process optimisation. But the real paradigm shift is end-to-end value chain orchestration - stitching data seamlessly from raw material procurement to shop-floor operations, and all the way out to logistics and real-time customer demand. Technologies like Industrial IoT, digital twins and computer vision will become as commonplace as ERP is today. Not to replace people, but to help engineers solve harder problems faster.
The manufacturing leaders of tomorrow won't be judged by plant capacity in metric tonnes alone, but by data velocity - how fast their systems learn and adjust.
"The factories that learn the fastest will lead the industry the longest."
Beyond Pilots: Three Areas Where AI Moves the Needle
Anoop Kubba: Moving from vision to reality - a common complaint is that AI shines in a pilot sandbox but struggles to deliver ROI at scale. If you look at actual steel operations today, which three areas are creating genuine, uncompromised business value?
Prakash Tripathi: If I look at what moves the needle right now, I would prioritise three pillars: production optimisation, predictive maintenance and energy management.
First, production optimisation. Balancing furnace inputs, refining schedules and rolling mill throughput is an incredibly complex, multi-variable problem. AI handles these scheduling equations in real time - maximising yield, optimising raw material usage and cutting operational waste.
Second, predictive maintenance. In a continuous-process industry like steelmaking, an unexpected outage on a critical asset doesn't just halt a line; it cascades through the entire plant. Moving from reactive or calendar-based maintenance to data-driven prediction lets our teams intervene before components fail - and even a few hours of avoided downtime has a significant business impact.
Third, energy management. Energy is one of our largest recurring cost lines. AI models continuously analyse consumption patterns and recommend process adjustments that save meaningful energy without touching product quality.
What connects all three is not the technology - it is better operational decision-making. AI delivers value only when it directly targets these massive cost and yield drivers.
"In manufacturing, better decisions lead to better performance."
Inside Shyam Steel's Transformation
Anoop Kubba: Let's talk about Shyam Steel's own path. Could you pull back the curtain on the core digital and AI initiatives your team has prioritised - what worked, and what surprised you along the way?
Prakash Tripathi: Our journey has been strictly anchored to business outcomes, never technology for its own sake. We started by modernising our core business applications to unlock operational efficiency and agility. Once the core was robust, we fortified our cloud architecture and built strong cybersecurity guardrails to create a secure, scalable digital ecosystem.
With that infrastructure in place, we moved into data visibility - integrating disparate operations into unified dashboards and analytics for faster decision-making. That laid the runway for our AI-driven initiatives in predictive maintenance, quality analytics and intelligent automation, with data and AI now supporting operational planning and business decisions across the enterprise.
The biggest surprise? Realising that the real measure of success is never the sophistication of the technology you deploy. It is how cleanly that technology integrates into the workflows of the people on the plant floor.
"The real measure of success is not the number of technologies we implement, but the business value they create."
Powering Smart Steel: Breaking the IT-OT Silo
One of the strongest themes running through the conversation is the convergence of information technology and operational technology - getting legacy machines, modern sensors and enterprise applications to speak the same language in real time.
Prakash Tripathi: We describe it as powering smart steel through connected intelligence. You cannot have an isolated IT team at headquarters and an isolated OT team running control systems at the plant. We have systematically tied these worlds together using Industrial IoT - pulling high-frequency sensor data from shop-floor machinery into real-time analytics engines.
That gives us end-to-end operational visibility. When a mill manager can see instantly how a minor calibration adjustment alters fulfilment metrics or raw material consumption, the IT-OT silo is broken. Data becomes the single source of truth driving the factory floor - and our vision is a connected steel plant where it drives every decision, lifting productivity, quality and agility together.
From Prediction to Self-Optimising Operations
Anoop Kubba: Predictive maintenance is easily the most talked-about AI application in heavy industry, yet so many proofs-of-concept stay stuck in the lab. What separates the programmes that scale - and how do we evolve toward truly self-optimising operations?
Prakash Tripathi: It fails when people assume it is an IT project. A machine learning model is useless if the data feeding it is spotty, or if maintenance crews don't trust the alerts and ignore them. Success requires reliable asset connectivity, quality data and tight cultural alignment between the engineers building the models and the mechanics running the plant.
Looking ahead, predicting failures is just the baseline. The exciting horizon is moving from prediction to AI-assisted, self-optimising operations - systems that don't merely alert a human that a threshold is being crossed, but recommend or initiate the best course of action, continuously monitoring and learning. The long-term vision is intelligent plants that minimise downtime and optimise performance, with human expertise firmly at the centre of governance.
0 to 1 Is Technology. 1 to 100 Is Transformation.
Anoop Kubba: That distinction between the technology working and the business adopting it brings us to scaling. What fundamentally changes when you take an AI use case from 0 to 1 versus from 1 to 100 across plants and functions?
Prakash Tripathi: Going from 0 to 1 is a technology challenge - you are proving an algorithm works under controlled conditions. Going from 1 to 100 is a business transformation.
When you scale, everything changes. Your architecture must handle enterprise-wide data pipelines. Your governance must define clearly who owns data accuracy across plants. Leadership commitment has to be sustained, and IT, OT and business teams must genuinely collaborate. Most importantly, the culture has to change - you actively upskill the workforce so people run confidently alongside these tools. AI scales successfully only when it stops being a special project and becomes a quiet, standard part of daily operations.
"Moving from 0 to 1 is a technology challenge. Moving from 1 to 100 is a business transformation."
The Connected Factory Runs on Trusted Data
Anoop Kubba: To power all of this, the underlying data has to be pristine. Many CIOs argue that AI success has little to do with picking the fanciest model and everything to do with trusted operational data. Has that been your experience?
Prakash Tripathi: Absolutely. AI is an amplifier, not a magic wand. Feed it poor, disconnected data, and it will simply amplify those inefficiencies at a faster rate.
An AI-ready manufacturing organisation requires a rock-solid data foundation: Industrial IoT, robust master data management, scalable enterprise architecture and real-time visibility across plant operations. When data flows cleanly across IT and OT systems, AI models suddenly deliver accurate, actionable insight. For manufacturers starting out, investing in data maturity will often return more than investing in another AI model.
"Data is not just the foundation of smart manufacturing - it is its biggest competitive advantage."
India's Opportunity to Leapfrog
Anoop Kubba: Let's broaden the lens. Between Make in India, the infrastructure boom and global supply chain shifts, our manufacturing story is evolving rapidly. How should Indian manufacturers view digital transformation compared with global legacy players?
Prakash Tripathi: Indian manufacturers hold a genuine, structural advantage: we are building our digital capabilities at precisely the moment when AI, cloud and Industry 4.0 technologies have matured. Unlike many global counterparts burdened with decade-old legacy systems, we can move straight into modern, connected manufacturing models.
Our challenges are not about the availability of technology. They are about accelerating adoption, closing the workforce skills gap, improving data maturity and engineering resilient, technology-driven supply chains. If we pair our manufacturing grit with digital intelligence, India won't just catch up to global standards - we will help define them.
"India has the opportunity to leapfrog, not just catch up. The winners will be those who combine manufacturing excellence with digital intelligence."
Cybersecurity Has Become a Production Issue
Anoop Kubba: As factories become hyper-connected, security boundaries dissolve, thereby expanding the attack surface. This makes cybersecurity an operational risk, not an IT checklist. How do you ensure cybersecurity keeps pace without slowing innovation or operations?
Prakash Tripathi: Cybersecurity has to be baked into the blueprint of every digital project from day zero. At Shyam Steel, our approach is to secure business continuity, not just data. As IT and OT networks merge, we operate a zero-trust security architecture with continuous monitoring, IT-OT security integration, regular risk assessments and deep cyber awareness across the organisation - including the factory floor.
An attack on an industrial environment can halt production lines and endanger people, not just leak information. In connected manufacturing, operational resilience and cyber resilience have effectively become the same discipline: protecting production, people and business continuity.
Three Years Out: When AI Becomes Invisible
Anoop Kubba: Looking into the crystal ball - if we sit down for this exact conversation three years from now, what will surprise us most? What becomes standard, and what gets exposed as hype?
Prakash Tripathi: The genuine surprise will be how invisible AI becomes - so deeply embedded in everyday manufacturing that we stop calling it out as a separate initiative. Intelligent decision support, autonomous process optimisation, connected operations and AI-assisted quality control will simply be how a modern steel plant runs.
What gets debunked is the flashy, isolated AI pilot. The industry will look back and realise that standalone tools built on messy data foundations were a distraction. The real winners will be the organisations that did the hard, unglamorous work - fixing data integration, upgrading infrastructure and building a disciplined culture of execution. AI alone does not transform a business. Strong data, skilled people and disciplined execution do.
Three Principles for Future-Ready Manufacturers
Anoop Kubba: Finally - many leaders across India are navigating the awkward middle ground between digital experimentation and enterprise-wide scaling. What three guiding principles would you leave them with?
Prakash Tripathi: If I had to distil our entire journey into three principles for building a future-ready organisation, they would be these.
1. Anchor everything to the business outcome. Never let technology lead the discussion. Every digital or AI initiative must solve a tangible, measurable operational challenge or cost driver.
2. Fortify your digital bedrock first. Invest in data quality, connected systems and zero-trust cybersecurity before attempting to scale complex AI.
3. Bring your people along. Technology fails the moment the workforce rejects it. Upskill your engineers and floor teams, build their trust in data, and make them co-authors of the journey.
At the end of the day, competitive advantage will not belong to the manufacturers who bought the most expensive software. It will belong to those who used technology to make sharper, faster, better decisions.
"The most successful manufacturers won't be those with the most technology - they'll be the ones who use technology to make better decisions, every single day."
Published by HT Digital Content Services with permission from TechCircle.