From monitoring to an AI-native operating intelligence layer.

BESS Intelligence Platform

• Fleet and site intelligence across SOC, SOH, usable capacity, RTE, availability and asset status.
• Capacity degradation, recoverable vs. permanent capacity loss, and weak-component detection.
• Cell/rack voltage, temperature and resistance analytics with peer-based anomaly detection.
• Warranty-risk monitoring and operating-envelope visibility.
• Safety precursor analytics: abnormal thermal rise, rack divergence, HVAC degradation and correlated risk signals.

AI Equipment Intelligence Foundation

• Expected-vs-actual behaviour models shared across Solar, Wind and BESS.
• Equipment health scoring, peer benchmarking, deterioration detection and recurrence analysis.
• Common event, loss-attribution and root-cause-analysis engines.
• Energy-impact ranking, so the platform prioritises what matters operationally.

AI Data Reliability

• Telemetry health score: stale/frozen sensor, timestamp and communication-quality detection.
• Sensor-validation and data-confidence indicators attached to every AI finding.
• Traceable KPI provenance, so operators can see what a recommendation is based on.

Operations Chatbot Expansion

• Asset-level queries, anomaly summaries and historical comparisons.
• Conversational access to "what changed?", "what is abnormal?" and "what requires attention?" workflows.

Wind Intelligence Platform

• Turbine availability, expected-vs-actual power curves and wind-speed-normalised performance.
• Turbine-to-turbine benchmarking and persistent underperformance detection.
• Yaw misalignment and pitch-system analytics.
• Condition intelligence for gearbox, bearings, generator, converter and transformer where telemetry permits.
• CMS/vibration integration where available, recurring fault analysis and predictive health indicators.
• Wake-adjusted performance analytics (beta).

Forecast-to-Schedule Intelligence

• Day-ahead and intraday Solar and Wind forecasts with block-level outputs.
• Probabilistic P10/P50/P90 confidence bands with continuous forecast-error learning.
• BESS SOC, MW and MWh availability forecasting.
• Declared schedule vs. expected generation, likely deviation, and schedule-risk alerts.
• Estimated DSM exposure and forecast-driven recommendations for the Indian market.

Solar Autonomous O&M — Phase I

• Tracker commanded-vs-actual angle, stuck-tracker and intermittent-actuator detection.
• Tracker-specific loss calculation, inspection prioritisation and actuator degradation analytics.
• AI soiling estimation and cleaning ROI calculation.
• Recommended cleaning date based on expected recoverable energy, not a fixed schedule.
• Robotic cleaning integration: schedule, coverage, machine health, missed areas and post-cleaning recovery verification.

BESS Intelligence 2.0

• Warranty-envelope tracking across throughput, SOC window, temperature exposure, cycling and RTE.
• Degradation trajectory and Remaining Useful Life models.
• Thermal-risk precursor analytics and HVAC-performance deterioration.

nxtGrid Operations Copilot 2.0

• Portfolio-wide conversational intelligence across SCADA, alarms, weather, CMMS, equipment history, forecasts, schedules and maintenance records.
• Explains why generation dropped, which equipment to inspect first, and which issues are costing the most.
• Compares asset deterioration across 7/30/90-day windows and identifies similar historical failures.
• Generates evidence-backed RCA with confidence, recommended action and quantified energy/revenue impact.

AI Morning Operations Brief

• Daily summary of yesterday's losses, abnormal assets, today's forecast, priority work, schedule risk and BESS opportunities.

AI Incident Investigator

For trips and sudden MW drops

• Reconstructs the event timeline, correlates alarms, telemetry and weather, compares historical patterns and recommends the next inspection or action.

Digital Twin — Phase I

• Data-driven operational twins for solar inverters, transformers and trackers; wind turbines; and BESS PCS, containers and racks.
• Continuously compares expected state vs. actual state, quantifies deviation and tracks deterioration.
• Starts with operational/statistical twins; adds physics models selectively where they materially improve accuracy.

Predictive Maintenance & AI Work Planning

• 7/30/90-day failure-risk scoring and expected intervention windows.
• Remaining Useful Life models where sufficient telemetry exists.
• Prioritises work using technical severity × recoverable generation × financial impact × weather × spare availability.
• Creates CMMS work orders, recommends inspection priority, and measures post-maintenance recovery.

Hybrid Energy Optimiser

• Joint optimisation using solar forecast, wind forecast, BESS SOC, degradation cost, PPA/schedule obligations, export limits, weather and plant availability.
• Recommends BESS charge/discharge schedules for renewable firming and schedule adherence.
• Identifies curtailment-capture, peak-shifting and export-optimisation opportunities.
• Quantifies expected operational and financial benefit before recommending action.

Commercial Intelligence

• Translates every material technical issue into energy impact, revenue impact, recoverable energy and recoverable revenue.
• Revenue-at-risk views for derating, tracker faults, soiling, wind underperformance and BESS efficiency/degradation.
• Portfolio ranking by recoverable commercial value, not only technical severity.

Contract & Warranty Intelligence

• Tracks PPA obligations, plant availability requirements, OEM warranties, LTSA commitments, and BESS degradation/RTE guarantees.
• Automatically surfaces emerging compliance or warranty risks and the operating behaviour contributing to them.

Autonomous Operations — Phase I

AI-generated. Human-approved.

• Detect issue → diagnose likely cause → calculate energy and financial impact → rank urgency and recommend action.
• Prepare a work order or control recommendation; monitor resolution and verify recovered performance.
• Learn from maintenance outcomes and operator feedback.