Neuromorphic Readout for Hadron Calorimeters

2025-02-18

AI Summary

High-energy physics experiments require real-time, ultra-low-power characterization of hadronic showers in sampling calorimeters—particularly for future high-luminosity colliders. Method: This work proposes an end-to-end spatiotemporal modeling framework based on spiking neural networks (SNNs), directly processing time-resolved scintillation signals from PbWO₄ crystals to jointly estimate deposited energy, photon emission location, and spatial shower distribution—fully preserving topological shower structure without conventional volume segmentation. It introduces the first SNN-based hadronic shower model and a custom III–V semiconductor photonic integrated circuit architecture optimized for neuromorphic hardware, enabling microsecond-scale inference and picojoule-level energy efficiency. Results: Evaluated on simulated data, the method achieves high-precision parameter estimation with inference latency <1 μs and power consumption three orders of magnitude lower than GPU-based approaches, establishing a scalable neuromorphic sensing paradigm for next-generation collider experiments.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Hardware-aware ML
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

We simulate hadrons impinging on a homogeneous lead-tungstate (PbWO4) calorimeter to investigate how the resulting light yield and its temporal structure, as detected by an array of light-sensitive sensors, can be processed by a neuromorphic computing system. Our model encodes temporal photon distributions as spike trains and employs a fully connected spiking neural network to estimate the total deposited energy, as well as the position and spatial distribution of the light emissions within the sensitive material. The extracted primitives offer valuable topological information about the shower development in the material, achieved without requiring a segmentation of the active medium. A potential nanophotonic implementation using III-V semiconductor nanowires is discussed. It can be both fast and energy efficient.

Problem

• Simulate hadron impacts on calorimeter
• Process light yield with neuromorphic system
• Estimate energy and position using neural network

Innovation

• Neuromorphic computing system
• Spiking neural network
• Nanophotonic III-V semiconductors