Inside CEAT’s digital core: Debashish Roy on turning AI from enabler to business driver
Not every digital transformation story starts in a boardroom. CEAT’s began on the factory floor — when plant operators stopped waiting for the digital team to push AI adoption, and started asking, “Where else can we apply this?”
That shift in mindset is at the heart of a remarkable journey that has taken CEAT’s Chennai plant into the World Economic Forum’s Global Lighthouse Network — a recognition reserved for the world’s most advanced manufacturing facilities. Behind that milestone is a deliberate, KPI-driven approach to AI: platforms like IntelliAIMix optimizing compound mixing in real time, computer vision systems catching defects before they become costly rejections, and predictive maintenance replacing guesswork with machine-health data.
We sat down with Debashish Roy, Chief Digital Transformation Officer at CEAT, to understand how AI moved from a support function to a core business driver — cutting energy-related conversion costs by 20%, accelerating new product development by 10–15%, and even laying early groundwork in quantum computing with IIT partnerships.
What emerges is a clear philosophy: technology adoption isn’t the goal — measurable, scalable business impact is. Here’s how CEAT is building that future, one data-driven decision at a time.
Some edited excerpts:
What are some of the key technology initiatives that have made a huge impact?
At CEAT, our digital transformation journey has focused on solving real manufacturing and business challenges rather than implementing technology for its own sake. Some of the most impactful initiatives include AI-driven process optimisation in manufacturing, computer vision-based quality inspection, predictive maintenance systems, AI-enabled demand forecasting, and advanced analytics for supply chain and logistics optimization.
One of the standout initiatives has been IntelliAIMix, our AI-powered compound mixing optimisation platform. It helped improve productivity, reduce energy consumption, and most importantly reduce process variability by making the mixing process more stable and predictable.
We have also deployed AI-based visual inspection systems across multiple manufacturing stages to improve quality consistency and reduce dependency on manual inspection. In supply chain operations, our Export Container Optimisation Tool significantly improved container utilization and reduced Order-to-Dispatch turnaround time.
Beyond individual projects, the biggest impact has come from building a digital culture where plant teams, operators, engineers, and business functions increasingly use data to drive decision-making.
When you look back at CEAT’s AI journey so far, was there a defining moment where digital stopped being an enabler and became a core business driver?
Yes, there was a clear transition point. Initially, digital initiatives were viewed as support systems focused on reporting, monitoring, or incremental improvements. The mindset shifted when AI-led solutions started directly impacting core business KPIs such as productivity, throughput, quality consistency, energy cost, and customer responsiveness.
The defining moment came when plant teams themselves began asking, “Where else can we apply AI?” rather than the digital team pushing adoption. Solutions like IntelliAIMix demonstrated that AI could influence critical manufacturing outcomes safely and reliably in real production environments.
Once leadership and operations teams saw measurable business impact at scale — including improvements in conversion costs, faster decision cycles, and higher process stability — digital became a strategic business capability rather than just a technology layer.
CEAT’s Chennai plant being inducted into the World Economic Forum Global Lighthouse Network is a major milestone. Beyond technology adoption, what operational or cultural shifts truly set the plant apart?
The Lighthouse recognition was not only about adopting advanced technologies; it was about transforming the way the organisation thinks and operates.
One of the biggest differentiators was the strong collaboration between operations, digital, maintenance, quality, and business teams. Technology initiatives were deeply integrated into day-to-day manufacturing rather than being run as isolated pilot projects.
The plant also embraced a culture of experimentation and continuous improvement. Teams became more open to data-driven decision-making and were willing to challenge traditional operating methods. Operators and frontline teams played an active role in identifying problems and validating digital solutions.
Another important shift was the move from reactive operations to predictive and proactive operations. Instead of waiting for breakdowns or quality deviations, teams started using real-time insights to prevent issues before they occurred.
The focus was always on scalable business impact — improving productivity, quality, sustainability, and agility simultaneously.
Computer vision and predictive maintenance are often talked about abstractly. How have these capabilities changed day-to-day decision-making for plant managers and operators?
The biggest change is that decisions are now faster, more proactive, and increasingly data-backed.
With computer vision systems, operators no longer rely only on manual observation or periodic sampling. Defects can now be detected in near real-time with far greater consistency. This enables immediate corrective action, reducing rejection rates and preventing quality escapes.
Predictive maintenance has similarly transformed equipment management. Earlier, maintenance was largely reactive or schedule-based. Today, machine health indicators allow teams to identify abnormal behavior before a failure occurs. Plant managers can prioritize interventions based on risk and machine condition rather than fixed schedules.
This changes the operational mindset significantly. Teams spend less time firefighting and more time optimising performance. Operators also gain greater confidence because decisions are supported by live data and predictive insights rather than assumptions alone.
Supply chains today are defined by volatility. How has AI-driven demand forecasting changed CEAT’s ability to respond to uncertainty rather than just plan for efficiency?
Traditional forecasting models perform reasonably well in stable conditions but struggle during sudden market fluctuations. AI-driven forecasting allows us to incorporate a much wider set of variables and continuously adapt predictions as conditions change.
The biggest advantage is agility. Instead of planning only for efficiency, we can now respond faster to shifts in demand patterns, regional variations, seasonality, and changing customer behavior.
Improved forecasting also enhances alignment between production planning, inventory management, procurement, and logistics. This helps reduce stock imbalances while maintaining service levels. Shifts us from “firefighting shortages” to “planning buffers”.In today’s environment, resilience and responsiveness are becoming just as important as operational efficiency, and AI is helping us move toward more adaptive supply chain planning.
The Export Container Optimisation Tool has significantly reduced Order-to-Dispatch TAT. What does this tell us about the hidden inefficiencies AI can unlock in legacy processes?
One key learning is that many inefficiencies exist not because people lack expertise, but because legacy processes involve complex decision combinations that are difficult to optimise manually at scale.
Container planning traditionally depended heavily on experience, static rules, and manual coordination. AI helped identify optimisation opportunities across loading patterns, space utilization, dispatch prioritization, and planning sequences that were previously difficult to visualize holistically.
What is interesting is that these inefficiencies are often invisible because teams become accustomed to operating within existing constraints. AI helps expose these hidden optimisation layers by evaluating thousands of possible combinations quickly and consistently.
This demonstrates that even mature manufacturing and logistics processes still hold significant untapped value when approached with advanced analytics and optimisation techniques.
Cutting NPD time-to-market by 10–15% is no small feat in manufacturing. How is AI reshaping the collaboration between R&D, design, and production teams at CEAT?
AI is helping break traditional functional silos by creating a more connected and data-driven development ecosystem.
Earlier, product development cycles often involved sequential workflows where learnings moved slowly between R&D, testing, manufacturing, and quality teams. Today, digital tools and AI models allow faster simulation, data sharing, and validation across functions.
For example, historical production data, material behavior, quality trends, and test results can now be analyzed together to support faster design decisions and reduce trial iterations.
Manufacturing teams also become involved earlier in the development cycle, helping ensure that designs are optimised not only for performance but also for manufacturability and scalability.
The result is faster collaboration, shorter validation cycles, and quicker movement from concept to commercialization.
Your work with IITs on quantum computing for compound discovery is forward-looking. How do you decide when to invest in frontier technologies versus scaling proven AI use cases?
Our approach is to maintain a balanced innovation portfolio.
Proven AI use cases that deliver measurable operational value are prioritized for scaling because they directly improve business performance and help build organisational confidence in digital transformation.
At the same time, manufacturing is evolving rapidly, and we believe it is important to explore frontier technologies early enough to understand their future potential. Collaborations with IITs on areas like quantum computing allow us to build foundational knowledge and experiment in controlled environments without disrupting core operations.
The decision framework typically considers three factors:
Business relevance and long-term strategic value
Technology maturity and implementation feasibility
Capability-building potential for the organisation
Not every frontier experiment leads to immediate deployment, but these initiatives help prepare the organization for the next wave of industrial innovation.
CEAT has achieved a 20% reduction in energy-related conversion costs using AI. How do you frame sustainability initiatives internally—as responsibility, efficiency, or competitive advantage?
We see sustainability as a combination of all three.
First, it is a responsibility because manufacturing organisations must contribute toward reducing environmental impact and improving resource efficiency.
Second, sustainability initiatives often drive strong operational efficiency. Energy optimisation, waste reduction, and process stability directly improve conversion costs and productivity.
But increasingly, sustainability is also becoming a competitive advantage. Customers, partners, and global markets are placing greater emphasis on responsible manufacturing practices. Organizations that can produce efficiently with lower environmental impact will be better positioned for long-term growth.
What is encouraging is that AI enables sustainability and business performance to improve together rather than forcing a trade-off between the two.
Looking ahead, what are some of the exciting technology initiatives planned? Can you briefly take us through the same and the impact expected?
Looking ahead, our focus is on scaling connected intelligence across the manufacturing value chain.
Some of the key areas include:
* Expanding AI-driven autonomous process optimisation across more manufacturing operations
* Strengthening computer vision systems for real-time quality assurance
* Building advanced digital twins for simulation and operational decision support
* Increasing predictive and prescriptive maintenance capabilities
* Enhancing supply chain intelligence using AI-driven planning and optimisation
* Accelerating GenAI adoption for knowledge management, troubleshooting, and engineering productivity
* Exploring advanced material science applications through partnerships with academic institutions
We are also investing heavily in capability development so that digital innovation becomes embedded within operational teams and not limited to specialized digital functions.
The larger vision is to create highly adaptive, intelligent, and sustainable manufacturing ecosystems where decisions are increasingly proactive, connected, and data-driven across the enterprise.