ATLAS Kalman/Geometry framework

1
ATLAS Kalman/Geometry framework

• Similarities and differences of ATLAS/STAR
• Main features of the environment
• What can be of interest to us
• How to proceed

2
Purpose

• Year-2 needs re-thinking
• Increase efficiencies for primaries/secondaries
• Algorithms/sequence
• Tools
• Integration
• Gain from others experience
• STAR/ALICE/ATLAS (BABAR)

3
Similarities/Differences

• ATLAS3 pixel layers/4 doublesided SSS/TRT (no z
  info)
• 30 X0 at eta0 (5-10-10-5)
• Environment not ROOT-based. STAFlike now, GAUDI
• but well integrated/interfaced
• High pt physics (Tracker cut 500MeV-1GeV)
• but tracking tasks very similar
• Tracking approaches are different
• but tracking can start in TPC or SVT
• Use of geometry abstraction / database-like
  features
• Use of Kalman for filtering/smoothing too
• with extensions to event vertex and primary track
  fitting
• C/OO implementation

4
Similarities/Differences (cont.)

• Use of global/local coordinate systems
• propagationglobalspace points
• Kalman filterlocalclusters/errors
• Flexible track model
• Uniform B Helix
• Non-uniform Runge Kutta
• Tracks become fatter as more info is
  accumulated
• Flexible Helix class with operations/tools
• Electron/muon model is more important to them
• but not an issue

5
What can this environment do for us?

• We can use this environment as a lab to test
  ideas approaches
• It is easier to first test-adopt the whole thing
  and then replace particular modules than break
  functionality from the beginning
• It is better to have a choice/options than not!
• We can learn a lot in many areas by just playing
  with this framework
• Year-2 tracking RD group can use it as a seed
• It appears to be a better environment than the
  one we have in place now

6
How to proceed?

• Pavel can set-up the STAR geometry in the
  environment
• needs some response simulator/fast can be
  quickly implemented. He says 2-weeks will do.
• Igor needs to set-up tracking env./parameters
• 2-3 weeks. Not clear what his involvement will
  be. He might need assistance from STAR (Who?)
• STAR test drives the environment/optimization
• Unknown time-scale/availability/who?
• My feeling is that we should give it a shot! -)