AI infrastructure in Action: Think Digital Act Hybrid

Time of India

6 hours ago

Do you know that Indian factories now pour around ₹5.7 trillion every year—about US $69 billion—just into keeping their machines running? World-Bank data put India's 2023 manufacturing output at US $461 billion, and maintenance-economics studies show that routine upkeep typically consumes about 15 percent of production cost even in well-run plants.
The drain deepens whenever equipment fails without warning. A nationwide reliability survey found that 88 percent of industrial sites suffer at least one unplanned outage every month, and every lost hour costs the average plant roughly ₹7 million in scrap, overtime and idle labour. Pharmaceutical facilities feel the pain most sharply: recent research tracking disruption across life-science manufacturers shows downtime incidents lasting eight hours or more and losses that can top £5 million—about ₹52 crore—per hour once sterility has been breached.
The good news is that the numbers flip when factories move from calendar-based greasing to data-driven care. A cross-industry study of predictive-maintenance programmes reports about a 25 percent cut in maintenance spend and a 10–20 percent boost in equipment availability after plants fitted vibration, temperature or energy sensors and began using machine-learning models to flag anomalies early . Multiply those averages across India's manufacturing ledger and more than a trillion rupees a year could stay on the shop-floor instead of vanishing into unplanned downtime.
For decision makers in manufacturing industry, the route to those gains is straightforward. Begin by mapping which utilities / equipment / asset have the greatest impact on production integrity and uptime, then put continuous, always-on guardrails around that short list first. Feed the resulting data into a secure, vendor-neutral platform so operations, quality and IT share one source of truth; let anomaly alerts open work orders automatically; wrap the flow in zero-trust policies. Field evidence shows every rupee invested in this disciplined approach returns three to five rupees in avoided downtime while simultaneously hardening the plant's cyber posture. Several Indian manufacturers have already moved from pilot to production with exactly this playbook—most notably
Dr. Reddy's Laboratories
, which chose Kyndryl as its partner to make the vision real.
This is the logic behind the partnership announced in April 2025 between
Kyndryl and Dr. Reddy's Laboratories.
Under the agreement, Kyndryl is deploying its open integration platform, Kyndryl Bridge, to monitor every data-centre workload and every production asset across the pharma major's global network. The goal is a Zero-Touch operations model that uses AI to predict failures, trigger auto-remediation and cut manual interventions by about 60 percent, all from a single analytics dashboard that also simplifies compliance reporting.
Where traditional industrial-automation companies focus on selling the drives, controllers and sensors themselves, Kyndryl lives in the connective tissue: it ingests raw analogue loops alongside modern digital tags, normalises them into secure MQTT or OPC UA streams, applies zero-trust policies so a IoT device can talk to the cloud without inviting ransomware, and keeps the entire hybrid stack patched and audited 24 × 7.
In short, plants that wait for machines to fail pay twice—once for emergency repairs and again for lost production. Plants that 'sense analogue, think digital and act hybrid,' with services like Kyndryl's stitching old and new together, turn maintenance from a cost centre into a fountain of uptime, yield and margin. In a market where a single afternoon of sterile downtime can erase a quarter's profit, adopting Industry 4.0/5.0 maintenance is no longer an experiment; it is a balance-sheet imperative.
The author is
Kaustubh Ramchandra Purohit, Customer Technology Advisor, Kyndryl IndiaNote: This article is a part of ETCIO'S Brand Connect Initiative.