GEST: AI-Driven Energy Optimization for University Campuses

The Problem: Energy Inefficiency in Academic Settings

University campuses are energy-intensive ecosystems with diverse, dynamic loads: lecture halls peak at 120 kW during scheduled classes, dormitories draw a steady 80 kW overnight, and labs introduce transient spikes up to 50 kW. Legacy infrastructure (often 40% of HVAC systems pre-1990) leads to 25-30% energy loss through inefficiencies. Rising costs (5% annual increase) and carbon reduction mandates (50% by 2030) demand a solution. Green Energy Sustainable Technology (GEST), developed by GreenTouch, addresses this through integrated AI, IoT, and materials engineering, targeting a 35% reduction in energy waste and 200-ton annual CO₂ cuts.

System Definition: What is GEST?

GEST is an energy management platform for universities, combining machine learning, IoT sensor networks, and advanced energy storage to optimize campus-wide power usage. It forecasts demand, monitors real-time consumption, stores energy efficiently, harvests ambient power, and integrates renewables. Key components include:

• Load Prediction: LSTM neural networks for 95% accurate demand forecasting.

• Monitoring: 300+ IoT sensors for real-time data acquisition.

• Storage: SiC-enhanced phase change materials (PCM) for thermal energy retention.

• Harvesting: Pb(Zr,Ti)O₃ piezoelectric transducers for ambient energy capture.

• Renewables: Perovskite solar cells for on-site power generation.

Technical Implementation

GEST operates through a multi-tiered architecture:

• AI Forecasting: LSTM models, trained on 12-month datasets (e.g., 120 kW lecture hall peaks, 80 kW dorm loads), predict demand with a 2 kW mean absolute error, enabling 20 kW peak shaving during high-demand periods (e.g., 8 a.m. to 2 p.m.).

• IoT Network: Sensors (1 Hz sampling) monitor 300+ nodes, capturing voltage, current, and temperature. Data streams via MQTT to a cloud controller, which adjusts loads—e.g., reducing lighting by 5 kW or HVAC by 8 kW based on occupancy (detected via CO₂ sensors at 600 ppm threshold).

• Thermal Storage: SiC nanoparticle-doped PCM (0.8 W/m·K conductivity) stores 50 kJ per 1-inch cube, capturing off-peak heat (e.g., 2 a.m.) and releasing it during peak cooling (e.g., 3 p.m.), improving HVAC COP from 3.0 to 3.3.

• Piezoelectric Harvesting: Pb(Zr,Ti)O₃ transducers in high-traffic zones (0.1 MPa/step) generate 5W/m², powering sensor nodes.

• Solar Integration: Perovskite cells (1.5 eV bandgap, 15% efficiency) on windows produce 5W/m², maintaining 70% light transmission, supporting daytime loads.

Performance Metrics and Differentiation

GEST achieves measurable outcomes:

• Efficiency: 35% reduction in losses (150 MWh/year saved, $15,000 at $0.10/kWh for a mid-sized campus).

• Cost Savings: 20% lower operational costs via peak shaving ($5,000/month demand charge reduction).

• Sustainability: 200-ton CO₂ reduction/year, equivalent to removing 40 cars.

• Stability: Grid-interactive algorithms maintain voltage within ±5% during peaks.

Unlike static systems, GEST adapts via continuous learning, optimizing for campus-specific patterns (e.g., exam week surges) and integrating renewables seamlessly.

Target Applications and Collaboration

GEST is built for universities aiming to reduce energy costs and emissions while modernizing infrastructure. It also serves as a research platform for AI, IoT, smart grids, and renewable energy integration. GreenTouch seeks collaboration with academic and industry partners to advance GEST’s deployment and scalability.