How an Indian Pipe Maker is Quietly Building AI-Driven Factory Floor

Ashirvad by Aliaxis, the Bengaluru-based manufacturer of PVC and CPVC piping systems and a unit of Belgium-based Aliaxis with revenues of $778 million in FY24, had already laid much of its digital groundwork. The question now was where to apply AI, and how.

Chief Information Officer Vipin Rustagi says the deployment has been selective for a reason. "AI is an abused word these days, and I say that as someone who is genuinely invested in making it work," he tells AIM. "We have been very deliberate about where we deploy it, and that selectivity is intentional."

The Bengaluru-based company's focus landed on three areas: predictive maintenance, analytics-based demand forecasting, and anomaly detection for quality control. 

This reflects a broader logic common to manufacturers attempting AI adoption—pick use cases where the business impact is visible, and where the data already exists.

From Reactive to Proactive

Manufacturing decisions have traditionally depended on periodic reviews, manual monitoring, and responding to problems. "You waited for something to go wrong, and then you addressed it," he remarks.

The shift to IoT-enabled monitoring has changed the underlying rhythm of plant operations. Real-time data from machines now feeds into anomaly detection systems that can surface potential issues before they become failures. "The shift from reactive to proactive is not a small thing on the shop floor," Rustagi says. "It changes how decisions get made and how people operate."

This move is part of a pattern across Indian manufacturing. AI adoption in manufacturing in India jumped from 8% in FY23 to 22% in FY24, according to TeamLease Digital, reflecting a sharp rise in sector-wide integration. On the shop floor, predictive maintenance uses sensor data to anticipate equipment failures and can reduce downtime by up to 30%, per SAP.

Predictive maintenance accounted for 35–40% of the AI-in-manufacturing market in 2025, according to MarketsandMarkets, as industries move to reduce operational downtime and enhance equipment reliability. Ashirvad's deployment sits squarely within that trend.

The Technology Stack

Ashirvad's AI capabilities sit on top of a layered enterprise architecture. SAP ERP forms the operational core, with Snowflake serving as a data lake that consolidates supply chain and operational data. Salesforce handles customer relationship management, with both dealer and plumber-facing applications built on top of it. Microsoft Copilot is deployed within Office 365, and Salesforce Einstein is used for commercial analytics.

"We have built middleware and API layers specifically to ensure data flows without interruption across systems and decision-making platforms," Rustagi says. "Interoperability is what makes the data actually useful."

This architecture reflects a common challenge in the sector. Manufacturers face several hurdles in scaling AI adoption, including integrating AI with legacy industrial systems, improving data readiness, and modernising operational infrastructure.

Measuring the Return

To determine return on investment in AI, Ashirvad tracks reductions in unplanned downtime, improvements in forecasting accuracy, productivity gains, process efficiency, quality improvement,s and resource utilisation. 

"AI and automation investments are best assessed on their ability to generate measurable operational value and build long-term competitive advantage," he says. "If you cannot connect the investment to a business outcome, you should ask harder questions before you make it."

Anomaly detection has strengthened quality control by identifying variations that conventional inspection would not catch. Rustagi acknowledges these are incremental rather than dramatic gains. 

The Harder Problem

For manufacturers navigating AI rollouts, the bottleneck is not the technology itself. "The harder challenge is ensuring that your organisation is compatible with the innovation you are bringing in," he notes.

Plant teams in manufacturing environments prioritise stability. Uptime is the metric that matters most on the shop floor, and any technology initiative that threatens it faces resistance. "The job of IT leadership is to build integration pathways that respect that priority while progressively connecting the shop floor to the broader data ecosystem," Rustagi says.

Poor data integrity in legacy systems and the complexities of deploying AI across diverse manufacturing environments remain factors that may hinder market growth, a challenge that Ashirvad has attempted to address through phased implementation and governance frameworks rather than wholesale transformation.

Ashirvad by Aliaxis itself is in a period of expansion. The 28-year-old company plans to establish a new greenfield manufacturing facility in Hyderabad, targeted for commissioning by August 2026, and is upgrading its Durgapur plant with a compounding facility to enhance production capacity and automation. The company has set a goal of achieving $1.1 billion in revenue from India by 2030.

As those plants come online, the AI and data infrastructure being built today will be tested at a greater scale. For Rustagi, the foundation—not the headline use case—is what matters. "The deeper AI play becomes possible only after you have got this foundation solid."

India's government has positioned AI as a foundational pillar in the country's development ambitions, with a focus on building an inclusive AI ecosystem that empowers enterprises while preparing the workforce for the future. For manufacturers like Ashirvad, that ambition is being tested not in policy documents, but on the factory floor, one use case at a time.