Clarification by Andrea:
ASI would not lead the SGS, but to finance it, proposing a team with an italian scientist to lead it.
LM: in the notes Andrea, corrected Paolo’s comments. The comments by Andreas should be moved to the end, as comments.
Review of the current organizational proposal, from:
we should identify purely eu SGS tasks and international tasks
the Data Centers will be fully european, with a clear responsibility
AZ points out that the SGS is not only a computer center, but an organization to produce products
The SGS In EU will have the responsibility to deliver some product, tobe defined.
An EU SGS is then recommended
In our view, the responsibility will be given to WGs, composed by researchers all over the collaboration
The eu infrastructure will be responsible for assuring that EU scientists have access to data
Poalo reminds the structure defined in the Litebird concepts of operations and data utilization document:
AZ agrees that the definition of the products is a mission issue
But the development of algorithms must be organized. You can call it SGS or something else, still is needed.
RS: the definition of SGS and Data Center is critical here.
Andrea presents some slides:
we need to prepare the framework to analyse the data.
SGS EU PO Leads
SGS scientist: flowdown of scientific requirements, verifies that the products are of the expected quality
SGS technical elements
IOT coordinator: coordinates the operations of the instruments. In EU we are responsible of the MHFT. PM until commissioning. Later the responsibility of maintaining the instrument is passed to the SGS, in the person of the IOT manager.
To be agreed the interface with LFT.
The PO refers to higher level:
- global SGS coordination (at mission level, in US and Japan)
- report to global governance (a science team, ot the corresponding LB structire)
Under the SGS PO, there are the data centers
- computer and resources made available for the analysis
- algorithms are developed transversal to DC
- one assumption is that L2 and L3 are connected to the LiteBIRD SGS PO Leads,
- there are national responsibilities in level 2-3
- as you mention three different SGS (JP, US, EU) should coordinate
- how the US will coordinate is not our business. We need to define the interfaces
- In LB the idea is not to distribute SGS resp into nations, but to have a collaborative global organization
- Levels are not limited to Europe
- we are trying not to put competitive SGSs
- the goal is to have a working team on a specific algorithm, with people from all nations.
- no national responsibilities, but responsibilities on people in a WG
- no comparison of different pipelines
- our proposal is not incompatible
- the levels are spread in all regional SGSs
- in the division of tasks coordination is essential
- coordination can be implemented in several ways
- redundancy is not incompatible with a coordinate effort
- implementation will be a LB global decision
- if we have separated SGS each will use its resources, which is dangerous and counterproductive
- eu is in a peculiar position, and coordination will be particularly relevant.
- we are assuming that there are necessary interaction among the SGS
- task sharing of SGS activities
- very complicated structure
- in EU we have many voices, and we need infrastructure coordination
- when it come to Data analysis, we have the WG responsible for products, with rotation in the leadership
- the scope of the exercise it to try organize the work in europe
- organization is not easy, if we don;t organize now, we will discuss forever
- Otherwise we risk to have a disordered effort
- IOT is also key for MHFT
- first step is to define the structure to coordinate the EU contribution
- given Euclid experience, we need to act asap.
- there is an other motivation for coordination
- do we want EU to count 1 or 1/n?
- we need to avoid a fragmented situation
- about AZ comment: we all agree we need coordination
- we propose a coordination at international level, not eu level
- computing facilities with specific eu coordination
- IOT TBD
- for the rest, it should be a global coordination
- european organization is part of the global organization
- we need to define an EU organization asap, to be used as a clue to ask for a global SGS organization
- We have a proposal on the table clear
- we don’t want to be driven by political issues. This si also the view of Masashi.
- He wants an efficient structure, with motivated and skilled people, rather than a structure based on national requests and political reasons.
- I agree with this, but we need to have a balance between reality and science requirements
- National flags in the MHFT PO are legitimate, as a result of a compromise. And are not based only on competences. Those things count.
- We have competences, such as Andrea or Bouchet
- we can coordinate as Europe
- as we are doing for the PO
- If we fragment, we risk acting in a disordered way
- For example, what Oslo does in his Data Center, should be organized at EU level, and not left to their choice.
- these activities can be coordinated by JSGs
- some organization is needed for implementation of JSGs guidelines
- if we go continental way, EU will count 1/3.
- while we have half ot he members. We may have more role if we
- we can have an overall leading role if we get organized
- the description by AZ is nice at global level
- we can also place a bet for the global SGS, Europe has the expertise to do that.
- we have the competence
- we need financial support
- we need to convince Eu agencies to finance.
- we are trying to maximize the project result
- we need to coordinate resources, students, data centers, funding. We agree this is absolutely needed and requires a lot of effort.
- we need to agree if strategy is to coordinate at eu level first
- we need to make a plan of implementation, in order to have support from agencies
- we are talking of a service to the community
- we have to agree in Europe. As France took the MHFT, we (italy with ASI support) can take charge of the SGS
- in the interest of Eu data analysis experts we should coordinate.
- ASI is very welcome to support the global effort
- MHFT was given to europe for technical reasons
- The SGS can be easily shared
- MHFT has been assigned to CNES,
- then CNES organized the PO and asked and got PM, PI, Calib. scientist
- agency need some level of control of what they fund. This will be the case also if ASI funds SGS
- MHFT was defined by the collaboration before assignment to an agency and a PO
- we can look for a key role for ASI support in the global organization
- the collaboration needs to define the tasks.
- Do we agree we need to coordinate at EU level?
- I think that if we coordinate we get stronger and can better capitalize
- discussion is not converging. We need to change attitude
LM presents some slides, prepared to clarify my mind. Discussed with FP at some point, Shared with a few other French people. No global discussion.
PN: ASI is already paying for IMO as a service, without formal task-sharing
LM: I agree. This can evolve with more IT role in SGS
PN: Why only Data Analysis aspects under “international collaboration”? MHFT should also go under the “international collaboration” color.
LM: we try to map responsibilities here. CNES is responsible for delivery of MHFT
PN: some of the lower line can be responsibility of an agency (eg the IMO, the IOT)
As for the MHFT, the DA tasks need a leading agency, on top of contribution from all the partners.
LM: we don’t want to split responsibilities.
PN: we think assigning responsibilities is useful, and not only to the data centers Infrastructure, but also to the task. Agency pays for an infrastructure, but also wants to have some control of the process.
FP: the bottom line should include a much more complex proposal, similar to the top line. With contribution and possibly formal responsibilities not into “collaboration international”
LM: possible for IOT and IMO. more difficult for DA.
many discussions here are not going to the point
Main points for convergence:
- coordination at eu level has to take place or not?
- at least at the level of infrastructure, we agree
- also at the SGS level? as a maximization of the contribution at the international level?
- My feeling is that we need this level of coordination too.
- we all agree about the need of coordination of infrastructure
- two layers:
- international level
- EU level, primary business of the ESC
- implications for the governance, as we are doing for the MHFT PO
- Ludo to send slides. They include some proposal to be discussed next time.
- Paolo to share AZ slides