Enlil —
Unrolled 8-Step LSTM Controller
Enlil is the FCOG facility mechroom predictive controller. Named for the Sumerian god of wind, breath, and atmosphere — the fundamental forces an HVAC system manages. It uses an 8-timestep unrolled LSTM to predict equipment state transitions from 57 sensor inputs spanning the entire mechanical room: supply/discharge air temperatures, water temperatures, pressures, valve positions, VFD speeds, and current draw across boilers, pumps, and air handlers.
Abstract
Background
Enlil is the FCOG facility mechroom predictive controller. Named for the Sumerian god of wind, breath, and atmosphere — the fundamental forces an HVAC system manages. It uses an 8-timestep unrolled LSTM to predict equipment state transitions from 57 sensor inputs spanning the entire mechanical room: supply/discharge air temperatures, water temperatures, pressures, valve positions, VFD speeds, and current draw across boilers, pumps, and air handlers.
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)), Enlil achieves 418 FPS (hw_only) with 1.488 ms hardware latency at 0.83 W average power draw.
Conclusion
Enlil 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 prediction application.
| Model | Enlil |
| Domain | HVAC Prediction |
| Architecture | Unrolled 8-Step LSTM Controller |
| 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
Enlil is the FCOG facility mechroom predictive controller. Named for the Sumerian god of wind, breath, and atmosphere — the fundamental forces an HVAC system manages. It uses an 8-timestep unrolled LSTM to predict equipment state transitions from 57 sensor inputs spanning the entire mechanical room: supply/discharge air temperatures, water temperatures, pressures, valve positions, VFD speeds, and current draw across boilers, pumps, and air handlers.
Network I/O
Input: 8-step sensor sequence [1, 1, 8, 57]. Output: next-state prediction scalar.
Architecture
Unrolled 8-Step LSTM Controller
8-timestep unrolled LSTM with input_size=57, hidden_size=57. The recurrence is fully graph-flattened into sequential feedforward operations: 8 copies of the LSTM cell (4 gate projections × 2 weight matrices per timestep = 64 total matrix multiplications). Each timestep: input gate, forget gate, cell candidate, and output gate computed via separate Linear projections (no chunk/slice ops). Final hidden state feeds a Linear(57→1) prediction head. Compiled via DFC 3.33.1 with dead-layer-removal patched for recurrent compatibility.
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) | 418.10 |
| FPS (streaming) | 418.11 |
| HW Latency | 1.488000 ms |
| Power (streaming avg) | 0.83400 W |
| Power (streaming max) | 0.83800 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.