Towards 6D Tracking: A Study Of Using Fast-Timing For Measuring Track Position, Time, And Angles
Abstract
Next-generation particle trackers will incorporate precision timing with resolutions of tens of picoseconds. Using two-dimensional TCAD simulations of Low-Gain Avalanche Diodes (LGADs) with a 25 μm pitch and 75 μm thickness, we show that tilted tracks traversing multiple pixels produce systematic inter-pixel timing differences of hundreds of picoseconds. These differences must be corrected for in cluster hit time reconstruction, but they also encode the track angle. We derive an analytic linear model relating inter-pixel timing differences to the track angle, and determine the sign of the angle, which the cluster shape cannot provide, using an amplitude-weighted method. For tracks crossing at least three pixels, this enables single-layer angle reconstruction with a resolution of 2∘−3∘ and, after a simple numerical inversion, a bias below 1∘. The dominant limitation arises from large Landau fluctuations: non-uniform charge deposition along the track induces inter-pixel crosstalk that distorts the timing signals. Affected tracks can be identified and rejected using the total cluster charge. Neural networks that additionally exploit nonlinear correlations and pixel charge improve the resolution to approximately 1.7∘, indicating that the computationally inexpensive linear model captures most of the available angular information. These results provide a path toward 6D tracking, in which each detector layer measures track position, time, and angles.
Related
- Measuring the displacement of the center of pressure from an instrumented track while running: a comparative study
- From Hits to Tracks: A BERT-based Tracking Model for Track Reconstruction in Drift Chambers
- Using Line Segment Tracking to enable ML-based Track Reconstruction in CMS
- Characterization and Active Control of Position-Dependent Timing Dynamics in Superconducting Strip Single-Photon Detectors
- Twisted Kagome Bilayers: High-Order Van Hove Singularities, Sublattice Interference, Magic Angles, and Possible Topology
BibTeX
@misc{turbiner2026towards,
title = {{Towards 6D Tracking: A Study Of Using Fast-Timing For Measuring Track Position, Time, And Angles}},
author = {Victor Turbiner and Elena Villhauer and Saveer Jain and Abhiraj Gupta and Michael Cardiff and Julie Segal and Christopher J. Kenney and Mark Horowitz and Ariel Schwartzman and Angelo Dragone},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2605.24736},
eprint = {2605.24736},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2605.24736},
}
RIS
TY - UNPB
TI - Towards 6D Tracking: A Study Of Using Fast-Timing For Measuring Track Position, Time, And Angles
AU - Victor Turbiner
AU - Elena Villhauer
AU - Saveer Jain
AU - Abhiraj Gupta
AU - Michael Cardiff
AU - Julie Segal
AU - Christopher J. Kenney
AU - Mark Horowitz
AU - Ariel Schwartzman
AU - Angelo Dragone
PY - 2026
DA - 2026/10/03
PB - arXiv
DO - 10.48550/arxiv.2605.24736
UR - https://arxiv.org/abs/2605.24736
AB - Next-generation particle trackers will incorporate precision timing with resolutions of tens of picoseconds. Using two-dimensional TCAD simulations of Low-Gain Avalanche Diodes (LGADs) with a 25 μm pitch and 75 μm thickness, we show that tilted tracks traversing multiple pixels produce systematic inter-pixel timing differences of hundreds of picoseconds. These differences must be corrected for in cluster hit time reconstruction, but they also encode the track angle. We derive an analytic linear model relating inter-pixel timing differences to the track angle, and determine the sign of the angle, which the cluster shape cannot provide, using an amplitude-weighted method. For tracks crossing at least three pixels, this enables single-layer angle reconstruction with a resolution of 2^∘−3^∘ and, after a simple numerical inversion, a bias below 1^∘. The dominant limitation arises from large Landau fluctuations: non-uniform charge deposition along the track induces inter-pixel crosstalk that distorts the timing signals. Affected tracks can be identified and rejected using the total cluster charge. Neural networks that additionally exploit nonlinear correlations and pixel charge improve the resolution to approximately 1.7^∘, indicating that the computationally inexpensive linear model captures most of the available angular information. These results provide a path toward 6D tracking, in which each detector layer measures track position, time, and angles.
ER -