When the Factory Floor Gets a Brain

In many factories across India, Southeast Asia, and parts of Europe, the management infrastructure follows a familiar pattern: paper logbooks by the assembly line, WhatsApp threads sharing shift updates, Excel spreadsheets compiled at the end of each cycle, and verbal handovers among crews. While data is present, it remains disconnected.

That gap between data generation and data use has become one of the more persistent operational problems in modern manufacturing. 

A May 2026 survey of over 600 US leaders by L2L, a leader in connected manufacturing operations, found that 90% of organisations are increasing their software budgets, while 74% remain caught in reporting delays that slow production. This is a consequence of what the research described as a "data paradox": rising digital investment that fails to reach the plant floor in any useful form.

It is into this environment that the Df-OS (Digital Factory Operating System) platform has positioned itself. 

The Problem It Is Addressing

"Before Df-OS, factory decision-making often relied on manual records, paper logbooks, Excel spreadsheets, WhatsApp messages, verbal updates, and end-of-shift reporting. By the time information reached line managers or plant heads, the opportunity to take corrective action had often already passed,” says Rajat Srivastav, Founder and Chief Executive Officer of Df-OS.

This is not an unusual account. Research from Inwedo, a digital manufacturing consultancy, found that 70% of data generated by manufacturing operations is never used, and that only 16% of manufacturing executives monitor production processes in real time. 

At firms that do collect data, the more common problem is fragmentation: production teams tracking inventory in one spreadsheet, finance in another, maintenance in a third, with no mechanism to reconcile them.

At a recent Manufacturing Execution Systems (MES) and Industry 4.0 summit in Porto, it was mentioned that MS Excel is still the most used MES in the world. The consequences include missed insights, inefficient processes, and a growing operational divide between manufacturers who have built connected systems and those still working from clipboards.

The Df-OS pitch is that replacing this infrastructure does not require lengthy ERP implementations or bespoke software at each site. The platform uses a no-code architecture, allowing factory processes to be mapped and digitised through configuration. According to the company, this approach has enabled deployment across 650 factories, spanning multiple industries and geographies.

Consistency across sites is maintained through standard process libraries, reusable templates, common data structures, and role-based workflows. "Every factory needs to manage production, quality, maintenance, manpower, safety, utilities, material movement, audits, and escalation processes. Df-OS is built around this process-first understanding of factory operations,” says Avinash Garg, President & Chief Business Officer.

The 2026 Deloitte Manufacturing Outlook found that 80% of manufacturing executives plan to allocate 20% or more of their improvement budgets to smart manufacturing initiatives over the next three years, and that the primary challenge is no longer technology selection but the scaling of digital systems from isolated pilots to enterprise-wide deployment.

The AI Layer: Vish.ai

Beyond process digitisation, Df-OS operates an intelligence module called Vish.ai, designed to analyse operational data across shifts, production lines, and time periods to identify patterns that standard reporting would not surface.

The company describes one case where the system detected a consistent production loss that had been attributed to machine breakdowns but was, on closer examination, caused by something different:

Vish.ai found that a recurring production loss was not caused by machine breakdowns, as initially assumed, but by a pattern of micro-delays during process handovers. “Individually, these delays appeared insignificant and were not being escalated through traditional review mechanisms," Srivastav tells AIM.

The finding was that material readiness and operator waiting time during shift handovers, in aggregate, were creating a production constraint. Because each individual delay was minor, no existing escalation threshold was triggered. 

The pattern only became visible when the AI examined data across shifts, lines, and timestamps simultaneously. Corrective actions were operational rather than technical: improved handover discipline, earlier material staging, and better visibility for line managers before losses accumulated.

"The value of Vish.ai goes beyond anomaly detection," Garg says. "It helps teams understand where hidden constraints exist, why they occur, and what actions should be taken."

Predictive Maintenance: The Financial Case

The area of predictive maintenance is where the financial case for industrial AI has been most thoroughly documented. Industry data suggests manufacturing downtime costs a median of $125,000 per hour, with high-precision industries reporting figures up to $1 million per hour. 

PwC research indicates that IoT-based predictive maintenance delivers roughly $7 in return for every $1 invested. The predictive maintenance market stood at $14.29 billion in 2025 and is growing at 28% annually, according to OxMaint.

For Df-OS, the return is framed in terms that extend beyond simple downtime reduction, says Garg, adding, "Equipment breakdowns can affect production schedules, manpower utilisation, energy efficiency, product quality, dispatch commitments, and customer service levels."

Metrics tracked by clients include reductions in unplanned downtime, improvements in MTBF (Mean Time Between Failures) and MTTR (Mean Time to Repair or Recovery), maintenance compliance levels, spare-part consumption trends, and — in more mature deployments — contribution loss avoided and throughput improvement. 

A 2025 ISG report on manufacturing industry services noted that AI-powered analytics and predictive maintenance are now "increasingly embedded into core workflows to reduce downtime and improve the accuracy of forecasts", having moved from experimentation into operational use.

Market Position

Df-OS operates in a competitive segment. Tracxn counts approximately 250 active competitors in the connected factory and digitisation space, including funded firms such as Steelhead Technologies and MAVERICK Technologies. Df-OS raised external funding of $4 million from investors like, Piper Serica, Amber, and Accel Atoms.

India's Production Linked Incentive scheme, which generated Rs 1.46 lakh crore in investments across fourteen sectors by August 2025, has pushed manufacturers to modernise — though costs remain a barrier. A comprehensive factory digitisation for a mid-size Indian facility typically runs Rs 50 lakhs to Rs 2 crore, with ROI realised within 12 to 18 months, according to eCorpIT.

The question facing Df-OS is whether operational gains can be consistently demonstrated across 650 varied factory environments, and whether a no-code model holds as manufacturing complexity increases.

The problem they are addressing, however, is not in dispute. As Srivastav states, factories do not transform simply by connecting machines. They transform when people, processes, and decisions become connected.