Strategy Library (60+)
Predictive pre-cooling / pre-heating, dynamic load balancing, and zone-level comfort playbooks tailored to climate and system type.
What is AI-Box™?
The Autonomous Technology for Building & Energy Operations
AI-Box™ ingests live building signals, synthesizes adaptive strategies, and closes the loop every 5–15 minutes while respecting on-prem or hybrid deployment policies.
Technology Architecture
From Field Signals to Autonomous Commands
A single appliance ties together legacy automation, data fabric, AI brains, and the command bus.
Edge Appliance
Industrial-grade hardware deployed on-prem or private cloud
Data Fabric
Normalizes BAS/BMS, EMS, PLC, IoT with millisecond time stamping
AI & GenAI Brain
Hybrid physics/ML models and self-learning loops
Control Command Bus
Issues verified instructions via Modbus, BACnet, OPC-UA, or REST
Core AI Engine
Four Pillars of the AI-Box™ Brain
Each module is versioned inside the model factory, so new building types or climates can be onboarded without rewriting the full stack.
GenAI Strategy Engine
Generative agents simulate thousands of “what-if” combinations of weather, occupancy, and tariffs to recommend the optimal control plan in minutes.
Self-Learning & Diagnostics
Anomaly clustering and root-cause analytics surface valve sticking, sensor drift, or control conflicts and push actionable tasks.
Security & Deployment Flexibility
Supports on-prem edge, dedicated cloud, or hybrid rollouts with zero-trust communications, RBAC, and full audit traceability.
Implementation Journey
Five-Stage Enablement Flow
Implementation is a technical practice with clearly defined timelines, certified A/B testing, and transparent KPIs.
01
Questionnaire & Feasibility
Collect floor plans, load profiles, and utility bills to estimate saving potential.
1–2 weeks02
Data Acquisition & Diagnostics
Deploy edge appliance, connect to BMS/EMS/IoT feeds, and build digital baselines.
3–4 weeks03
Letter of Intent & Report
Deliver diagnostic insights, KPI definitions, and recommended control strategies.
~1 week04
AI-Box Deployment & A/B Test
Compare baseline vs. strategy weeks under matched weather windows following GB/T standards.
4–6 weeks05
Continuous Savings Operations
Closed-loop tuning, periodic energy audits, and lifecycle support.
OngoingA/B Testing Methodology
- Match similar days (temperature ±5%, humidity ±10%) to compare manual operation vs. strategy mode.
- Apply GB/T 28750, GB/T 32045, GB/T 31349, GB/T 13234 references for savings calculations.
- Produce electrical / thermal / gas audit reports plus comfort dashboards.
15-40%
Saving band (real projects)
±3℃
Temperature control accuracy
5-15 min
Dynamic optimization cycle
Whitepaper
Deep Dive Into the Technology
Download the AI-Box™ technical whitepaper covering architecture, model library, A/B test methodology, cybersecurity hardening, and procurement checklists for engineering & sustainability teams.