Zephyrus —
Temporal Convolutional VAV Optimizer
Zephyrus optimizes variable air volume terminal box operation — damper positions, reheat valve commands, and minimum airflow setpoints — from zone-level sensor data. Named for the Greek god of the west wind. Reduces simultaneous heating and cooling (the most common source of commercial HVAC energy waste).
Abstract
Background
Zephyrus optimizes variable air volume terminal box operation — damper positions, reheat valve commands, and minimum airflow setpoints — from zone-level sensor data. Named for the Greek god of the west wind. Reduces simultaneous heating and cooling (the most common source of commercial HVAC energy waste).
Approach
The model was trained in AxonML (a pure-Rust deep learning framework) and compiled through the Hailo Dataflow Compiler (DFC 3.33.1) targeting Hailo-8 silicon. Post-training INT8 quantization was applied during the DFC compilation pass with production telemetry calibration data. The resulting Hailo Executable Format (HEF) binary executes on Hailo’s fixed-function dataflow architecture with deterministic latency and zero framework overhead at the edge.
Results
On production hardware (Hailo-8 M.2 (P/N: HM218B1C2FAE, S/N: HLDDM2A234600289)), Zephyrus achieves 10,484 FPS (hw_only) with 0.118 ms hardware latency at 1.29 W average power draw.
Conclusion
Zephyrus is production-ready as a single HEF binary deployed to edge devices with no external dependencies beyond the HailoRT vendor runtime. The model meets real-time latency requirements for its target hvac control application.
| Model | Zephyrus |
| Domain | HVAC Control |
| Architecture | Temporal Convolutional VAV Optimizer |
| Target silicon | Hailo-8 |
| Measured on | Hailo-8 M.2 (P/N: HM218B1C2FAE, S/N: HLDDM2A234600289) |
| DFC compiler | 3.33.1 |
| Framework | AxonML v0.6 (pure-Rust, CUDA + CPU backends) |
| Author | Andrew Jewell Sr. · ORCID 0009-0005-2158-7060 |
| Organization | AutomataNexus LLC · Fort Wayne, Indiana |
Executive overview
Zephyrus optimizes variable air volume terminal box operation — damper positions, reheat valve commands, and minimum airflow setpoints — from zone-level sensor data. Named for the Greek god of the west wind. Reduces simultaneous heating and cooling (the most common source of commercial HVAC energy waste).
Network I/O
Input: 10-channel VAV terminal telemetry. Output: damper position + reheat valve command.
Architecture
Temporal Convolutional VAV Optimizer
6-layer dilated TCN with 10 input channels: zone temperature, zone setpoint, zone CO2, supply air temperature (from AHU), discharge air temperature, damper position feedback, reheat valve position, occupancy sensor, time-of-day encoding, day-of-week encoding. Predicts optimal damper position and reheat valve command to minimize zone deviation while avoiding simultaneous heating/cooling conflicts.
Compilation constraints
All AxonML models targeting Hailo silicon are compiled under the fixed-function dataflow constraints: no dynamic control flow, no variable-length dimensions, all activations representable in INT8 after calibration, and no operations requiring dedicated softmax hardware (replaced with ReLU gating or depthwise convolution equivalents where necessary).
Silicon performance
Measured on production hardware via hailortcli benchmark with 5-second sustained inference. Device: Hailo-8 M.2 (P/N: HM218B1C2FAE, S/N: HLDDM2A234600289).
| Metric | Measured Value |
|---|---|
| FPS (hw_only) | 10,483.90 |
| FPS (streaming) | 10,483.90 |
| HW Latency | 0.118000 ms |
| Power (streaming avg) | 1.29400 W |
| Power (streaming max) | 1.29700 W |
| Power (idle) | 0.74979 W |
| Quantization | INT8 (post-training, DFC calibration) |
| DFC Compiler | 3.33.1 |
| HailoRT | 4.20.0 |
| Measured On | Hailo-8 M.2 (P/N: HM218B1C2FAE, S/N: HLDDM2A234600289) |
Deployment
Deployed as a single HEF binary. No ONNX runtime, TensorFlow Lite, or Python inference stack required at the edge.
| Target silicon | Hailo-8 |
| Measured on | Hailo-8 M.2 (P/N: HM218B1C2FAE, S/N: HLDDM2A234600289) |
| DFC compiler | 3.33.1 |
| Quantization | INT8 (post-training, production telemetry calibration) |
| Runtime | HailoRT (vendor runtime) |
| Edge platform | Raspberry Pi 5 + Hailo AI HAT+ (M.2 Key M) |
Deployment procedure
Copy the .hef binary to the target device. hailortcli run loads the HEF directly into the Hailo-8 dataflow engine over PCIe. Inference begins immediately with deterministic per-frame latency. No model conversion, graph optimization, or warmup phase required.
References
- Jewell, A. (2026). AxonML: A Pure-Rust Deep Learning Framework for Edge Inference. AutomataNexus LLC. Technical whitepaper.
- Hailo Technologies Ltd. (2024). Hailo Dataflow Compiler User Guide. DFC v3.33.1.
- Hailo Technologies Ltd. (2024). Hailo-8 Product Datasheet.
AutomataNexus LLC · Fort Wayne, Indiana · andrew.jewellsr@automatanexus.com
Andrew Jewell Sr. · ORCID 0009-0005-2158-7060
May 2026 · All rights reserved.