NPPS Leadership Team

[Brett Viren <bv AT bnl.gov>]

Hi Torre,

It is really hard to interpret this vision into my frame of reality but
what I do interpret is very negative.  Hopefully I'm not too blunt but
here is what I read:

- Shields up around CSI/SDCC in form of some abstract interface
  separating facility resources from experiment users.

- Interface to be defined by CSI/SDCC and experiments must figure if/how
  it can be used.

- Current BNLers who are AI/ML-adjacent suck at AI/ML, hire new ones and
  put them in CSI.

- Build some vauge automation to solve unspecified problems, but mostly
  facilities related ones.

I hope my interpretation is just plain wrong as this is a pretty bleak
picture.

What I want to see is something that hits these points:

- Deeper integration of CSI expertise and personnel into BNL
  experiments.  CSI meets the domains, not the domains bow to the gates
  of CSI.  CSI scientists work alongside domain scientists to explore
  new ways to apply AI/ML to solve domain problems.  (We are having some
  amount of success now with this approach!)

- Tailored interfaces between SDCC and BNL experiments.  While providing
  those, SDCC seeks and applies ways for reuse and automation and
  resources sharing across its customers.

- Lowered/zeroed bar to BNL domain scientists use of GPU and other AI/ML
  hardware to enable uplifting of skills and development of new AI/ML
  techniques and their application.  Ie, like RACF but with GPU/HPC.
  Sure, require some "proposal" for large production running but if I
  have to defend a proposal for a few GPUs, I'll buy them myself - or
  more likely drop my idea and work on the myriad of other things going
  on.

- Recognize that most of the effort in applying AI/ML is domain-specific
  data wrangling work.  Once in a form AI/ML codes can accept, the hard
  work is done and the play time begins.  Given the variety of needs,
  putting up interfaces to infrastructure will only exacerbate what
  domain scientists need to do in the wrangling.  Better to have CSI
  work with domain scientists to educate them in ways to do the data
  wrangling, where needed, or have CSI actually contribute effort
  directly to that wrangling.  In doing so, common interfaces that
  actually work the way the scientists do may be developed to improve
  future data wrangling.

- Tie this 5-year "autonomous" thing to some thing more useful and
  concrete.  Simply automating facilities so facilities can facilitate
  is a poor vision.  Maybe without naming names ATLAS's automated
  production error response system or DUNE self-healing DAQ ideas can be
  invoked.


-Brett.


Torre Wenaus <wenaus AT gmail.com> writes:

> About the lab-wide efforts/goals in AI/ML... Frank Alexander is
> renewing the effort to put something together and I was asked to
> add a HEP part to this googledoc (it's a copy, Frank restricted
> access to the original). My section follows one from Elke, and
> Jerome has been asked for input also. We were each asked to produce
> a version of the table at the top of the document. Comments
> appreciated.
>
> https://docs.google.com/document/d/1au5ywJU1gvfLhqWr-n1ol0dxUYtBxdZchwMb-oWxui4/edit?usp=sharing
>
>   Torre
>
> On Thu, Jan 7, 2021 at 1:41 PM Torre Wenaus <wenaus AT gmail.com>
> wrote:
>
>     Darn.
>    
>     Thanks, Brett! Added
>    
>     On Thu, Jan 7, 2021 at 1:36 PM Michael Begel <mbegel AT gmail.com>
>     wrote:
>    
>         Nope!
>        
>         Michael Begel
>
>         On Thu, Jan 7, 2021 at 1:32 PM Torre Wenaus <
>         wenaus AT gmail.com> wrote:
>        
>             Beats me! Do you know?
>            
>             On Thu, Jan 7, 2021 at 1:29 PM Michael Begel <
>             mbegel AT gmail.com> wrote:
>            
>                 One comment is that I don't see the connection
>                 between the work outlined and the *lab's* goals in
>                 AI/ML. What are the lab-wide efforts and goals?
>                
>                 Michael Begel
>
>                 On Thu, Jan 7, 2021 at 1:21 PM Torre Wenaus <
>                 wenaus AT gmail.com> wrote:
>                
>                     Draft slides are here, comments appreciated...
>                    
>                     
> https://docs.google.com/presentation/d/1f5g4IXFS2_crHVfwf-rsEMseHc31V_2cOxJrs_0t884/edit?usp=sharing
>                    
>                       Torre
>                    
>                     On Tue, Jan 5, 2021 at 12:26 PM Brett Viren <
>                     bv AT bnl.gov> wrote:
>                    
>                         Torre Wenaus via Phys-npps-mgmt-l <
>                         phys-npps-mgmt-l AT lists.bnl.gov>
>                         writes:
>                        
>                         > And I have Brett's LDRD.
>                        
>                         In addition to the "LS4GAN" LDRD which
>                         pushes on a "new" (really a
>                         variant on existing) ML technique, some
>                         established techniques have been
>                         applied by Haiwang Yu (EDG postdoc) on two
>                         problems:
>                        
>                         1) Improve the ionization signal
>                         reconstruction efficiency in LArTPC for
>                         particularly problematic track topology. 
>                         (Paper accepted to JINST).
>                        
>                         2) Neutrino vertex finding in a sea of
>                         cosmic muons in the LArTPC
>                         MicroBooNE detector on the surface.
>                        
>                         -Brett.
>
>                     --
>                     -- Torre Wenaus, BNL NPPS Group, ATLAS
>                     Experiment
>                     -- BNL 510A 1-222 | 631-681-7892 |  
>                     wenaus AT gmail.com | npps.bnl.gov | wenaus.com
>                     -- NPPS Mattermost room: 
>                     https://chat.sdcc.bnl.gov/npps/channels/town-square
>
>             --
>             -- Torre Wenaus, BNL NPPS Group, ATLAS Experiment
>             -- BNL 510A 1-222 | 631-681-7892 |  wenaus AT gmail.com | 
>             npps.bnl.gov | wenaus.com
>             -- NPPS Mattermost room: 
>             https://chat.sdcc.bnl.gov/npps/channels/town-square
>
>     --
>     -- Torre Wenaus, BNL NPPS Group, ATLAS Experiment
>     -- BNL 510A 1-222 | 631-681-7892 |  wenaus AT gmail.com | 
>     npps.bnl.gov | wenaus.com
>     -- NPPS Mattermost room: 
>     https://chat.sdcc.bnl.gov/npps/channels/town-square
>
> --
> -- Torre Wenaus, BNL NPPS Group, ATLAS Experiment
> -- BNL 510A 1-222 | 631-681-7892 |  wenaus AT gmail.com | npps.bnl.gov
>  | wenaus.com
> -- NPPS Mattermost room: 
> https://chat.sdcc.bnl.gov/npps/channels/town-square

Attachment: signature.asc
Description: PGP signature