Speaker
Description
To address the challenges of future collider experiment environments, machine learning (ML) in readout electronics can be leveraged for smart detector designs, enabling intelligent inference and data reduction at-source. Autoencoders offer a variety of benefits for front-end readout; an on-sensor encoder can perform efficient lossy data compression while simultaneously providing a latent space representation that can be used for anomaly detection. Results are presented from low-latency and resource-efficient autoencoders for front-end data processing in a futuristic silicon pixel detector, enabling a readout scheme that can provide combined capabilities of off-detector data reduction and real-time sensor defect monitoring. Together these results highlight the multi-faceted utility of autoencoder-based front-end readout schemes, and motivate their consideration in advanced detector designs.
