Installed Is Not Modernised: Why the Digital Systems Line Decides Grid Modernisation

Grid modernisation is often reported as a count of devices: meters installed, feeders sensed, substations automated. Recent evidence from India and North America points to a different bottleneck. The hardware is arriving, but the software, data and operating models that turn it into decisions are lagging, and the demand forecasts that justify the investment are themselves under scrutiny.

India: the meter is installed, the data line is not

As of 30 June 2026, India had installed 7.24 crore smart meters against 20.33 crore sanctioned under the Revamped Distribution Sector Scheme (RDSS), with a deadline of 31 March 2028, according to an analysis by Grid published on 18 August 2026. The same piece reports that the Ministry of Power attributes delays to procedure rather than technology: late request-for-proposal issuance and vendor empanelment, slow tendering and award, delays in signing Direct Debit Facility agreements, public-private partnership implementation challenges, and weak consumer engagement.

The more consequential finding is where the money goes. The analysis states that only about 3% of metering funds target the digital systems that convert raw meter data into billing, loss statements and work orders. It also reports that of ₹100.9 crore allocated for personnel training, only ₹23.36 crore had been spent by July 2026, with 14,853 employees trained. A DISCOM can therefore reach a high installation count while its billing, collections and energy-accounting processes still run on pre-meter assumptions.

North America: from pilots to platforms

The pattern is not unique to India. A February 2026 review of lessons from DistribuTECH argues that utilities have moved beyond conceptual grid modernisation and now face operational scalability and reliability questions, with experience drawn from advanced distribution management system (ADMS) and distributed energy resource management system (DERMS) deployments. Its central warning is that digital grid platforms are advancing faster than enterprise operating models, data governance and workforce capability. It also notes that vendor emphasis has shifted towards interoperability, modular architectures and demonstrable return, and that utilities now scrutinise scalability and total cost of ownership more rigorously than in the pilot era.

A concrete example is the Imperial Irrigation District in California. Local reporting from May 2026 describes a $36.7 million programme, with $18.3 million of federal funding and $18.4 million from the district, centred on an ADMS for more than 165,000 customers. The system is meant to consolidate several separate software programs into a single real-time platform, enabling faster fault detection and, in some cases, automated rerouting. No final completion date has been set, which is typical: the platform is a multi-year integration effort, not a product delivery.

Load forecasts are now part of the modernisation case

Grid investment is justified by expected demand, and that expectation is being questioned. Utility Dive reported on 23 September 2026 that more than $170 billion in AI data centre capacity has been blocked, withdrawn or stalled since January 2024, according to advisory firm Relae. Texas paused new data centre interconnections pending an audit of a 474 GW queue, more than five times the grid's recent peak of 90 GW; ERCOT's chairman called much of that load speculative. Grid Strategies found utility data centre forecasts submitted to FERC were roughly 40% above credible industry projections through 2030.

States are responding with large-load tariffs that include minimum contract terms, minimum billing demand, collateral, upfront payment for impact studies and exit fees. The relevance beyond the United States is methodological: network upgrades sized to unverified load requests risk becoming stranded cost.

What it means for utilities

For Indian DISCOMs, the lesson from the RDSS data is sequencing. Meter installation, SCADA and DMS works, and IT and OT upgrades are only valuable when the data pipeline, billing integration and field workflows are funded and staffed alongside them. A budget in which digital systems and training are small residual lines will tend to produce installed but unused assets.

The data centre experience suggests a second discipline: treat large-load requests as contracts, not forecasts. Where India expects significant new demand from data centres, tariff design that requires commitment before network build is a way to protect both the DISCOM and its consumers.

Finally, ADMS-class platforms should be planned as integration programmes. The IID example shows the value proposition, consolidation and faster restoration, but also that completion dates are uncertain until integration scope is understood.

Key takeaways

• India had 7.24 crore smart meters installed against 20.33 crore sanctioned at 30 June 2026; the RDSS deadline is 31 March 2028.
• Only about 3% of metering funds reportedly target the digital systems that make meter data usable; training spend lagged allocation.
• Operating models, data governance and skills are advancing more slowly than grid platforms.
• Large-load forecasts can overstate demand; commitment-based tariffs are an emerging safeguard.
• Modernisation is best measured by decisions enabled, not devices installed.

Sources

• India's Smart Meter Rollout: Why Data Matters Before March 2028 (Grid)
• Grid Modernization and AI in Utilities: Lessons from DistribuTECH
• IID secures $36.7 million investment to modernize power grid (NBC Palm Springs)
• Efforts to curb data center speculation gain ground across the US (Utility Dive)