Speaker
Description
Modern high-energy physics experiments generate large data rates, requiring fast and efficient online processing. Embedding machine–learning–based feature extraction directly in the front-end electronics of the detectors is a promising approach to reduce the amount of data to be transmitted. Field-Programmable Gate Arrays (FPGAs) offer a unique platform for such tasks due to their parallelism, reconfigurability, and low power consumption.
Within this project, we investigate Edge AI for feature extraction and data preprocessing directly on FPGAs. Using a workflow that combines knowledge distillation, pruning, and quantization, we compress deep neural networks into compact models that can run on SoC/FPGAs with minimal latency.
As a proof of concept, we implemented a multi-layer perceptron via hls4ml on an FPGA to perform pulse-shape discrimination for the COMPASS/AMBER ECAL2 calorimeter, achieving a sub-microsecond inference latency and competitive accuracy with minimal resource usage.
This study opens the door for further exploration of alternative network architectures, extended training datasets including a broader range of pulse types, and the feasibility of multiclass classification according to distinct signal morphologies. Such developments could expand the approach's applicability and inspire new strategies for integrating machine learning into high-rate data acquisition systems.
CERN group/ Experiment
AMBER experiment
| Working area | Area 2: Optimal AI deployment for Online Data Processing |
|---|---|
| Project goals | Deployment of FPGA based ML algorthims for data reduction in the AMBER ECAL2 electromagnetic calorimeter |
| Timeline | first proof-of-concept done; 1 year for first in-beam tests; 3 years for full deployment in experiment |
| Available person power | 0.3 FTE (fellow/PhD) |
| Additional person power request | 0.5 FTE (fellow) |
| Is this an already ongoing activity? | Yes |
| Indicative hardware resources needs | interactive gpu cluster for model training of "teacher" networks |