Every AI model in SenseCoreAI is documented in peer-reviewed research with permanent DOIs. You can read exactly how each model works — and why the results are trustworthy.
Energy forecasting means predicting how much electricity your building will consume in the future — hour by hour, day by day. Without a forecast, you are always reacting to problems after they happen. With one, you can plan budgets, shift loads to cheaper times, and prevent overspend before it occurs.
SenseCoreAI uses a Temporal Fusion Transformer (TFT) — a state-of-the-art deep learning model trained on 36.4 million half-hourly electricity readings from 1,112 real UK commercial buildings. It produces 24-hour ahead forecasts updated every 30 minutes, achieving a forecast error of just 10.31% MAPE — verified in peer-reviewed research.
Anomaly detection means automatically identifying when a meter, building, or piece of equipment is behaving unusually — consuming far more or less energy than expected. Without it, a faulty piece of equipment or a meter error can go undetected for months, running up thousands of pounds in unnecessary costs.
SenseCoreAI runs three independent anomaly detection methods simultaneously on every meter in your estate. Z-Score analysis flags statistical outliers. Interquartile Range (IQR) detection catches skewed distributions. Isolation Forest uses machine learning to identify genuinely anomalous patterns. When two or more methods agree, an alert is raised — reducing false positives while catching real issues.
Fuel poverty occurs when a household cannot afford to heat their home to a safe, healthy temperature. It affects 3.1 million households in England — most of them in social housing. The Social Housing (Regulation) Act 2023 and Warm Homes Plan require registered providers to identify and act on fuel poverty. The problem is that most vulnerable households never self-identify.
SenseCoreAI scores every property from 0 to 100 daily using five signals extracted from meter data alone — no surveys, no home visits, no tenant interaction required. Properties scoring above the critical threshold trigger automatic welfare referral alerts. The five signals are: consumption level, temporal distribution, seasonal variation, consumption volatility, and LSTM Autoencoder reconstruction error.
UK energy regulation places significant statutory obligations on public sector organisations. SECR (Streamlined Energy and Carbon Reporting) requires annual carbon reporting for large organisations. ESOS (Energy Savings Opportunity Scheme) requires four-yearly energy audits. ERIC is an annual NHS estates return. SHDF requires grant evidence for social housing retrofit. DSP Toolkit governs NHS data security. Each carries financial penalties for non-compliance.
SenseCoreAI generates each compliance report automatically from your meter data using the correct specification for each framework. SECR reports use current DEFRA Scope 1 and Scope 2 conversion factors. ERIC reports are mapped to NHS England's prescribed format. ESOS evidence packs include AI-generated savings analysis ready for Lead Assessor sign-off. All reports are exportable as PDF or structured data.
Knowing when a building is occupied — and when it is not — is fundamental to energy management. Heating an empty building wastes money. Failing to cool an overcrowded one wastes energy and creates risk. Traditionally, occupancy data required expensive sensor networks or manual logging. Most organisations simply do not have it.
SenseCoreAI infers occupancy patterns from electricity consumption data alone using Gaussian Mixture Models (GMM). By clustering consumption patterns over time, it identifies occupied and unoccupied periods without any physical sensors or hardware installation. The result is an occupancy profile for every building — updated continuously from existing meter data.
Scope 1 emissions are direct emissions from sources your organisation controls — gas boilers, on-site generators, fleet vehicles. Scope 2 emissions are indirect emissions from the electricity you purchase. Together they form the basis of SECR reporting and net zero target tracking. Most organisations calculate these manually once a year — often with errors and out-of-date conversion factors.
SenseCoreAI tracks Scope 1 and Scope 2 emissions continuously from your meter data using current DEFRA conversion factors that are updated automatically. Every reading is logged to an immutable audit table, creating a continuous carbon record that can be used for SECR filing, net zero progress reporting, and ESG disclosures. Reduction over time is tracked automatically.
Every feature in SenseCoreAI is accessible via a single authenticated REST API. Connect your meter data once and access forecasting, anomaly detection, FPRS, compliance reports, occupancy inference, and carbon data — all from the same endpoint.
30 minutes. Your own buildings. We walk through every capability live — forecasting, anomaly detection, FPRS, and compliance reports.