Level 1 Scouting Workshop

3 Sept 2023, 19:30 → 6 Sept 2023, 11:30 Europe/Zurich

Sala dei Congressi (Centro Congressi di Sappada )

Emilio Meschi (CERN), Marco Zanetti (Universita e INFN, Padova (IT))

After the Phase-2 upgrade in 2029, the CMS Level-1 trigger will include tracks from the L1 track trigger, full-granularity information from the barrel calorimeter fine grained primitives from the new endcap calorimeters, as well as high-resolution data from the muon systems. Phase-2 trigger physics objects will have accuracy comparable to offline for transverse momenta down to 2 GeV and pseudorapidities up to 2.5 
The L1 data scouting project, whose principle architecture is exposed in the L1 Phase-2 upgrade TDR, aims at collecting the L1 intermediate data from the trigger processors at the collision rate of 40 MHz, apply data reduction, re-calibration, and classification online, and make them available for specific physics analysis, possibly to also be carried out online.
The principle of the L1 data scouting was pioneered in Run-2 by collecting the intermediate data from the global muon trigger. A more complete demonstrator is being commissioned at this time for operation in Run-3, collecting data from the GMT, the Layer-2 calorimeter trigger, the BMTF, and the Global Trigger.
The aim of this workshop is to review the areas of physics that can potentially benefit from the  phase space the scouting approach gives access to, including opportunities provided by novel approaches such as ML techniques. The status and plans of the Phase-2 L1 data scouting project will be summarized, including alternatives to the baseline, in particular concerning the distributed online processing approaches, and the status and results from the Run-3 demonstrator will be presented. The desired outcome is a more in-depth definition of the physics case and a plan of work for simulation studies to be carried out in the years leading to Run-4, as well as possible analyses accessible with Run-3 data. 

Monday 4 September

1 Welcome and Introduction to the workshop

Speakers: Dr Emilio Meschi (CERN), Marco Zanetti (Universita e INFN, Padova (IT))

Intro.pdf

Sappada.pdf

Sappada.pptx

2 Current results from searches for exotic signatures

A review of the experimental results (mainly from CMS and ATLAS, but not only) on signatures with non-SM final states.

Speaker: Livia Soffi (INFN Roma I (IT))

L1-Scouting-EXO-Summary_v1.pdf

10:15 Coffee break

3 Where and how we can improve on exotic signatures

With focus on phase-2, what improvements could we expect thanks to the new trigger system and new detectors

Speaker: Andrea Rizzi (Universita & INFN Pisa (IT))

Rizzi-Sappada.pdf

4 Design and limitations of the 40 MHz scouting system for Run 3

The plans for the 40 MHz scouting system during Run 3, its ultimate design and limitations.

Speaker: Dinyar Rabady (CERN)

L1 scouting architecture and limitations.pdf

12:00 Lunch at Agriturismo Heidille

Wonderfull chalet at walking distance from the workshop venue (15 minutes)

5 Status of current 40 MHz scouting system: a technical overview

Technical description of the current infrastructure, both hardware and software. Path towards the ultimate implementation for Run3.

Speaker: Thomas Owen James (CERN)

sappada_2023_technical.pdf

6 Properties of the current scouting data: GMT Muons

The data we are gathering, its quality and features. Their usage as tool for monitoring and studying the L1 system.

Speaker: Sabrina Giorgetti (Universita e INFN, Padova (IT))

gmtmuons_L1Workshop_040923.pdf

gmtmuons_L1Workshop_040923.pptx

7 Properties of the current scouting data: Calorimetric Objects and BMTF primitives

The data we are gathering, its quality and features. Their usage as tool for monitoring and studying the L1 system.

Speaker: Mr Rocco Ardino (Universita e INFN, Padova (IT))

L1ScoutingW_CaloL2_rardino.pdf

8 Fast Machine Learning for online scouting studies: applications and implementations

Using machine learning to re-calibrate and classify L1 objects for improved online analysis potential. Implementation in FPGA with flexible high-level tools

Speaker: Thomas Owen James (CERN)

sappada_2023_ml_short.pdf

9 Scouting detector front-end data: the DT case

Phase 2 electronics being developed for the DT chambers allows reading out at 40 MHz the whole full-resolution detector data. The past and the current sector test at CMS is going to be presented as well as the next steps and possible developments.

Speaker: Andrea Triossi (Universita e INFN, Padova (IT))

40MHz_scouting_AT.pdf

16:00 Coffee break

10 Possible measurements with current scouting data

Are there physics measurement that can be performed with the legacy L1 trigger information?

Speaker: Marco Zanetti (Universita e INFN, Padova (IT))

ScoutingAtRun3.pdf

ScoutingAtRun3.pptx

11 Are there New Physics signatures that could have escaped so far?

A review of models whose signatures either could have escaped if not purposefully searched for or which due to the current way data are managed/processed/analyzed we have little sensitivity for

Speakers: Zhen Liu, Zhen Liu, Zhen Liu (University of Minnesota (US))

24_MissingExotics_CMS_L1_2023.pdf

12 40 MHz scouting for Luminosity monitoring

How the scouting could be used to monitor the Luminosity, which source is best suited, the biases that could affect the measurements.

Speaker: Cristina Oropeza Barrera (Universidad Iberoamericana (MX))

Scouting-for-Lumi_Scouting-Workshop_04-Sep-2023.pdf

19:40 Dinner at Agriturismo Zaine

https://goo.gl/maps/XjaPCcDR835xvFgZ6

Tuesday 5 September

13 Phase II L1 trigger objects and menu: the physics potential of a 40 MHz scouting system

Review of the expected performances of the Phase II L1 trigger reconstruction and L1 Physics Menu. How close the L1 reconstruction will it be to the offline one, and which physics potentials can have a 40 MHz scouting system with respect to the L1 Physics Menu

Speaker: Cristina Botta (University of Zurich (CH))

L1TObjects_and_Menu_ScoutingPotential.pdf

14 The 40 MHz scouting system for Phase II

How the current infrastructure will evolve and what its main figures will be.

Speaker: Dr Emilio Meschi (CERN)

DSPhase-2-EM-Sappada5September2023.pdf

11:00 Excursion

We propose a short and easy hike to Rifugio Calvi at Monte Peralba (350 m elevation, 45 minutes walk on a gravel road), where it will be possible to have lunch. Car pooling will be organised (20 minutes from workshop venue)

Alternatives include an easy walk along Piave river to Rifugio Piani del Cristo, or a visit to the old part of Sappada.

15 HLT Scouting overview: results, perspectives, limitations

Overview of the HLT scouting, the current infrastructure, the results obtained so far and those expected in the future. What are the main limitations and solutions to overcome them?

Speaker: Mario Masciovecchio (Univ. of California San Diego (US))

Mario-Masciovecchio_L1ScoutingWorkshop_05September2023.pdf

16 Computing models for online/offline analysis

Solutions should be found on how to manage and process the amount of scouting data. How shall the online analysis be organized? What shall be left for offline?

Speaker: Matteo Migliorini (Universita e INFN, Padova (IT))

L1ScoutingSappada_mm_v1.pdf

17 New Physics agnostic analyses: anomaly detection on scouting data

Triggerless data acquisition allows a partial removal of the biases with which we search for NP; 40 MHz scouting is a suitable place where to employ anomaly detection techniques (a.k.a. model-independent analysis). How shall that be practically organized? Can it be run online?

Speaker: Thea Aarrestad (ETH Zurich (CH))

AD_scouting.pdf

18 GNN based τ->3μ analysis

Speaker: Miaoyuan Liu (Purdue University (US))

tau3mu_L1ScoutingWorkshop2023.pdf

20:00 Dinner at Edelweiss Stube

Edelweiss Stube is an excellent restaurant at the beginning of the valley, not far from Bach Boutique Hotel.

Wednesday 6 September

19 The Enhanced Physics Reach of CMS with an enabled scouting system

How the physics potential of the experiment will benefit from a scouting system. A few examples and comparisons.

Speaker: Maurizio Pierini (CERN)

L1ScoutingWorkshop_Sep23.pdf

20 Wrap-up and Conclusions

Summary discussion. Definition of the to-do list and next steps

Speaker: Jacopo Pazzini (Università e INFN, Padova (IT))

40MHzScouting_wrapup.pdf