From: <dbarnes_at_email.protected>
Date: Mon, 5 Jul 2004 10:00:50 +1000

Aus-VO Access Grid Meeting
Tuesday 29 June 10:30

Minutes (DR)
Chair (DB)

Attendance
  David Barnes
  Rachel Webster
  Mathew O'Dowd
  Becky Brown
  Jon Smillie
  David Holder
  Robert Power
  Peter Lamb
  Markus Buchhorn
  Neil Killeen
  Malte Marquarding
  David Abramson
  Jagan Kommineni
  Radha Nandkumar
  Michael Drinkwater
  David Rohde

Apologies
  Sarah Maddison
  Anne Green

1. Welcome

David Barnes thanked David Rohde for organising the meeting, noted
that organisation was difficult, try to address the time of next
meeting in this meeting.

Hopefully this can be a forum for doing demonstrations of various
things we are all working on - although it will be important to check
what the local facilities provide - e.g. powerpoint.

2. Aus VO and APAC Astronomy grid project

APAC is involved in funding for grid projects, of which some money is
going towards VO projects. There is funding for 4 new positions
around the country. The grants are going to be divided up somewhat so
that Aus-VO focuses on the science side of things and the APAC focuses
on computing.

3. Projects at different institutions, and comments about e-research
    broadly.

MELBOURNE (DB)
--------------
Report: Project Scientist (DB) part funded on Aus-VO grant, continuing
with national and international coordination, and local management
issues. Matt O'Dowd appointed as VO & Gemini scientist, half Aus-VO
money, and will be starting work shortly on a Gemini instrument
pipeline. Casual money spent on Brett Beeson (to finish SkyCat, and
develop theory portal prototype), Michael Lancaster (to develop VOlume
and AusVOdownload utility) and Christine Chung (some work on
theory "pipelines", eg. Zeus-MP taking input from another theory code).

SWINBURNE (provided by SM pre-meeting)
--------------------------------------
Report: 2.5 years of money came from APAC/VPAC - programmer being
hired to do theory portal work - hope to have somoene on the ground by
the end of July. Theory codes to go into the portal already chosen -
work has begun preparing one of the codes - also working on ideads of
theory code metadata.
DB - requests info on Pulsar activites under Aus-VO 2004 grant.

MONASH (DA)
-----------
1. We have done some initial implementation work applying our GriddLeS
technology (www.csse.monash.edu.au/~davida/griddles) to a simple
astronomy pipeline built from legacy components (provided by David
Barnes). This is still in early stages, but the initial results are
positive - ie. we have coupled a number of codes using pipes. GriddLeS
makes it possible to couple legacy codes using either files or pipes,
and these can be distributed across grid resources. This approach makes
it possible to use legacy code (written in any language) without the
need to perform significant software modification.

2. Deployment of Nimrod family of tools
(www.csse.monash.edu.au/~davida/nimrod) continues both nationally and
internationally. Initial discussions with Rachel suggest there are some
interesting applications of one variant of Nimrod (Nimrod/OI -
optimization + interactive) to complex astronomy optimization tasks. We
are currently building some novel interfaces for Nimrod, including a web
service interface and a Matlab interface. There may be some interesting
applications of this technology to astronomy and I would welcome input
on this.

3. DA recently went on a quick tour of UK eScience centers in Cambridge,
Oxford, Southampton and Cardiff. It looks like there will be application
projects for Nimrod at all of these sites, over a range of science
disciplines. It is also likely that Cardiff will act as the primary
point of contact for Nimrod in the UK.

4. Yet to advertise new positions at Monash, but this will occur over
the next few months.

ANU (JS)
--------
The ANU VO efforts are materially supported by the APAC Grid Program,
yeilding one nominal EFT currently comprising Jon Smillie and Tom
Kobialka. The two main elements of the ANU VO work are (1) VO-enabling
legacy data-sets, and (2) supporting and grid-enabling the Australian
Gravity-wave research community. (1) is currently focussing on the
MACHO data-set, in particular the WCS augmentation of the MACHO image
archive, accompanied by a wholesale reformatting of the MACHO metadata
to yeild a legacy XML metadata set. (2) is directed towards rolling
out a Lightweight Data Replicator(LDR) infrastructure to support the
data analysis and management activities of the ACIGA project.

CSIRO ICT (PL)
--------------
* Using astronomy as a main motivator in eScience

* Establish an archive of the data history of the Australia Telescope
Compact Array with metadata describing the file contents, and inferred
metadata grouping files into observations with associations between
science
targets and calibration targets.

* Value-add to the archive by supplying Web Services (and later
Semantic Web Services) for to reduce archive data to more useful data
products. Currently have a Web Services pipeline based on AIPS++ scripts
written by ATNF (Tara Murphy & Neil Killeen) that goes form an SQL
fragment
describing the data required to an image downloaded to the client
machine
and displayed using ATNF's Remote Visualiastion System.

* Algorithms for efficient near-neighbours-within-radius as a selection
filter for catalog crossmatching - a plane-sweep algorithm that is doing
spatial join between billion and half-billion object databases in 3-5
hours.

QUEENSLAND (MD)
---------------
2003: 1-year UQ internal cross-discipline "research development grant
funded 1 computer science/astrophysics RA position. We developed working
machine learning software to match astro catalogues (see presentation by
David Rohde).

2004: Lief grant will continue funding of this position to produce a
prototype catalog matching service. We are looking for input on what the
requirements might be.

David Rohde will attend the NVO summer workshop in
September (DATES 12 -17 September)

We are talking to more people in computer science at UQ as well as
Dave Abel (CSIRO ITC, Canberra) about closer collaboration on the
matching problem.

Marcus Gallagher (Computer science) is on sabbatical at Edinburgh
and collaborating with Bob Mann and his group there.

UQ in general has many related activities in place, notably recent
appointment of Prof Bernard Pailthorpe (Advanced Computational Modelling
Centre) from Sydney. I'm not sure of the "corporate" response to the ARC
yet.

UNSW (provided by WC post-meeting)
----------------------------------
Funding from the 2004 "Australian Astronomy Grid" LIEF grant that has
been allocated to UNSW is being combined with its own cash contributions
plus funding from the "Gemini" LIEF grant to create a 2 year "Deputy
Gemini Scientist/VO-astronomer" position. The primary duties of the
person who fills this position are:

(i)assist the Australian Gemini Scientist in the day to day running of
the Australian Gemini Office,

(ii) provide support, particularly in the area of data reduction
software, to Gemini users within the wider Sydney area,

(iii)develop a data reduction pipeline for one of the Gemini
instruments, and

(iv)work on the integration of data being collected at Siding Spring
Observatory for the global ROTSE-III experiment into the VO
environment. The position was advertised in April with a closing date
of 3 May. Four applications were received, three of which were
interviewed for the position. An offer has been made and it is hoped
that the appointee will take up the position in September.
SYDNEY (from AG)
----------------
LIEF appointment sorted out, to continue work on MOST pipeline. AG
going on sabbatical for six months, so seeking interim management of
USyd end of Aus-VO.

ATNF (NK)
---------
VO Projects Progress
--------------------
Remote Visualization System (RVS) - presently two people (Anil Chandra and
Praveena Tokachichu; both ARC LIEF money) working on this. We plan to
deploy RVS to ATNF image archives by Sep 2004 and make it publicly
available (i.e. download for use in your own archive) by end 2004.

ATCA Archive online - the archive resides physically at CSIRO ICT Centre
in Canberra. The interface will be taken over by ATNF (Chris Owen)
and the archive will be publicly available (for file download)
by Sep 2004.

ATCA pipeline - Tara Murphy (CSIRO postdoc) is working on this. We
plan to deploy it to the ATCA archive in an automatic mode
with limited functionality (single pointing, continuum) by Sep
2004. It will be integrated with the ATCA archive interface.
RVS will be deployed on the image products of pipeline.

We will push into other modes of operation (spectral line,
mosaicing after this initial deployment). The pipeline will
eventually be used online at Narrabri and offline.

Web Services - a simple (aips++) quanta service has been made
and will be fully publicly available by Sep 2004. We are then
going to develop the more interesting (aips++) measures
(frame conversions e.g. LSRK -> TOPO) web service

New Hires
---------
We have another CSIRO postdoc to hire in the emerging science
area. I hope to have them onboard by end 2004. This person
will work in the area of automatic source charactersiation
and finding in images.

E-science
---------
CSIRO is going ahead with a project to upgrade the links to our
observatories. From this we are initiating an e-VLBI project to do VLBI
in real time with the Australian network.

3.5 ARC LIEF funding 2004

Rachel Webster sent a memo to all partners. ARC have changed rules
and require a more formal agreement. Every institute's Research
Office will need to sign the agreement, Becky will be in charge of
organising this.

Also require project statements from each group with what they plan to
do with funds. Gant charts also a good idea.

8. Will LIEF grant get rolled in to e-research. It may be possible to
   change the LIEF grant to get funding for multiple years.

Ah Chung Tsoi trying to get new money for a new ARC program in
e-Research.

At this point international trips and international linkages are not
funded by ARC.

Trips have been beneficial though, possibly it would be better to have
longer trips rather than people to spend three days in a plane for
three days working.

The initial proposal is for a one year pilot program.

It was suggested that exchanges between Australian institutions for
between 3 - 6 months may also be valuable. Particularly of benefit to
programmers working in small groups. David Barnes and David Abramson
mentioned that they were happy to host people.

Aus-VO Workshop 2004
Last 2 were very useful particularly international visitors.

Alex Szalay has offered to attend.
Other suggestions for guests
Robyn Allsman of NOAO ex ANU
Bob Mann at Edinburugh
Doug Tody
Ray Plante

Meeting Time: Dates for the Aus-VO meeting in November between
supercomputing 2004 and thanksgiving, to allow American visitors to
attend. DB to follow-up with potential visitors and determine dates.

Proposed Aus-VO meeting dates during week November 15 - 19.

Attendance at meetings
IVOA Interoperability meeting, September 27-29

IVOA small projects meeting, September 30 - October 1, 2004, Pune,
India.

ADASS XIV meeting, October 24-27, 2004, Pasadena, USA

Super Computing 2004, People from ?ANU said they will attend.

NVO Summer School, September 12-17, David Rohde will attend.

Next meeting
Early September
11 - 17 September David is at NVO summer school
Not the 9th but soon after would be good for Neil Killeen.
DB to follow up.

Proposed meeting date 10 September.

4. Feed back on the Aus-VO working groups, and WIKI pages.
Less than spectacular response so far.

- The topics are ok, it is just too low a priority
- Maybe we don't need it at this stage
- Critical mass problem, we need more people (and more people are coming
on board)
- AG Meetings are better because they force a deadline
- The head of groups are too busy to do it, we could either nominate
somebody to maintain it, or simply not use it.
- DB set it up with the hope that it will be a repository of
important information - failed in this respect so far - but DB still
thinks it is neccessary

- Mabye too many groups, for critical match. Resembles a building
  with lots of doors that need to be checked.

David Barnes said he will think about it and suggest what to do.

Presentation by David Rohde about Machine Learning and Catalogue
Matching

Presentation by Malte Marquarding about Remote Visualisation System.

Meeting Closes at 12.40.
Received on 2004-07-05 10:01:52