All Dots Connected - Preemptive feature search and online indexing for next-generation HEP experiments (#1081)
1 CERN, EP/CMD, Geneva, Genève, Switzerland
Several future HEP experiments under design or construction, including the HL-LHC upgrades, will feature massive high-precision trackers, extreme granularity calorimeters, some based on silicon sensors, and improved muon systems capable of high-efficiency identification. The use of fast optical links enables unprecedented data rates to fully exploit these detectors, however, in many instances, power/cooling infrastructure and the subsequent material budget force the choice of a two-stage data acquisition to limit the readout rates. “Intelligent” detectors have been proposed - along with corresponding fast hardware pattern-recognition engines - to overcome the inherent efficiency limitations of a two-level trigger system and optimize the use of available bandwidth. The amount of intelligence that can be put in the front-end is however limited by the harsh radiation environment. An overview of the different approaches to exploit these powerful detectors while performing data reduction at early stages, along with potential alternatives, including alternative readout schemes, will serve as an introduction to discuss new approaches to fully exploit their physics potential. In particular, we will focus on one hand on common practices for online data reduction and selection and how they can profit from techniques, new and old, which are nowadays ubiquitous in other fields of data science. On the other, we will discuss how we can make use of the massive amount of data at different resolution to preemptively identify interesting features, thus enabling an entirely new approach to data reduction based on real-time indexing of these features. We will conclude with a rapid overview of the relevant technologies available or anticipated on the market and a look ahead at possible future developments.
Keywords: Trigger, Tracking, Data Acquisition, Data Science, Calorimetry
Challenges and opportunities in data-intensive synchrotron-based imaging and microscopy (#1145)
1 Argonne National Laboratory, Lemont, Illinois, United States of America
As the sophistication of today's experiments grow at synchrotron light sources, collecting the most informative data has become greatly relevant, necessitating the development of methods and techniques that can provide good quality reconstructions from big data streams. Overcoming these challenges commonly requires developing better approximations of physical systems, and when these approximations are not available or too costly to compute, approaches based on machine learning can help in automating the process. In this talk, I will first give a broad overview of the status of imaging and microscopy applications, and then describe how existing big data, compressed sensing, and machine learning methods can be adopted to enable faster and reliable information extraction from complex measurement data. I will also highlight the need for an integration of hardware and software in building successful instruments of the future, especially after realization of the next-generation of x-ray sources providing orders of increased brilliance and coherence.
Keywords: imaging, microscopy, tomography