Paris May 10-11 summary

1) Organization
	The IPT (Integrated Product Team) organization was supported by all group. It was also supplemented by a "client/vendor" (or "customer/supplier") approach in which components required by some packages are provided by a unique "supplier" to (eventually multiple) "customers" for integration. "Customers" are ultimately responsible for the definition of the deliverables (although they clearly must be agreed upon with the "supplier" if we want to have one...), "suppliers" are responsible for the on-schedule supply.

2) Antenna
	The antenna groups agreed on a vertical division for the construction work. This means a single, commonly accepted, design, to be produced according to blueprints by whichever contractor any of the 3 partners may select. Each partner will thus provide as "in-kind" contribution 21 antennas. One will provide an extra antenna to match the total number of 64.
	
	There are worries that the antenna cost may be higher than currently budgeted

3) Receivers
	The receiver group proposes an integrated organisation following the "customer/supplier" model, in which common sub-components are only fabricated in one place. Major ensemble like warm optics, cryo system, individual cartridges will be integrated in "integration centers" (one for each ensemble). The receiver group also proposes a unique integration center for the overall integration of these major packages into a functional receiver. They recommend this integration center to ultimately move to Chile.
	Receiver costs have increased.

4) LO and Backend
	There was no big issue on that item. Several schemes of re-distributing the tasks were mentionned, but the details depend mostly on whether a complete photonic LO system 
can be build or not.

5) Correlator
	Costing of the 3 different approaches pursued by the 3 groups were presented.
	The total costs mentionned are
		Baseline correlator (US)	17 M$
		Future correlator	  (EU)	22 M$ (+ 20 % contingency ?)
		Enhanced correlator (Ja)	43 M$ (+0, -20 % contingency)
	Savings made possible by early decision on whether to build the Baseline correlator or only 1 quadrant amount to 3-10 M$, depending on decision date (8 M$ for the single quadrant option). 
	Induced computing costs were estimated to about 5M$, but only for an average doubling of the output data rate, while initial estimate suggest a factor 4 to 5 in average data rate increase from the baseline correlator to any of the other option.
Part of that cost is duplication of correlator software, which may have been accounted
twice.

6) System Engineering
	To be shared equally between the partners

7) Software
	A detailed proposal for a 40-40-20 division of effort was discussed. The cost of software increased significantly.
		
8) Science
	Scientific support needs to be increased, and shared equally by the three partners.
A true list of full-time ALMA scientists is urgently needed. 
The ACA feasibility remains unclear (can we calibrate it accurately enough ??) and would require an additional 6 months of effort (1 person full time, i.e. 2 astronomers). DSB vs
SSB issues were briefly discussed. 
	
EAEC Meeting
------------
	Following the Team leaders meeting, an EAEC meeting was held on Saturday 12, in Paris to attempt to evaluate the cost and organisation implications. The consensus on the antenna procurement model is a major progress. However, detailed tasks division remains 
difficult and depend on yet undecided options (e.g. photonics LO, or even more problematic correlator issue, or number of receiver bands to be produced). The scope of ALMA given the latest information should be re-assessed by ASAC for further progress.


MAJOR ASAC ISSUES:
------------------
	The re-costing excercise made while  looking at the 3-way project showed that we are unlikely to have sufficient confidence in the costs to allow for the ALMA Compact Array. 
Costs increased in the receiver and software area, and there is also a potential risk of antenna cost increase, given the status of the current prototypes. 

ASAC should recommend whether ACA studies should be continued or not.

	If ACA cannot be accomodated in the 3-way ALMA project (either because of cost or feasibility), it is essential to get ASAC recommendation on the additional bands respective priorities. For information, the current projected receiver costs are
(for ALMA only, no ACA)

	WVR		 9.0 M$
	Band 1	19.0 
	Band 3	16.3
	Band 4	16.5
	Band 6	20.6
	Band 7	19.4	(*)
	Band 8	23	(extrapolation from 7 & 9)
	Band 9	28.0	(**)
	Band 10	31	(extrapolation from 9)

(*) Probably only covers DSB, 4 GHz option, to be confirmed...
(**) As above ?