Until the final formal software architecture design (4 above) it is difficult
to accurately estimate the resources required for the WFCAM science archive.
In rough outline, and based on existing discussion documents and our
experience, we expect the following software `blocks', with staff resource
estimates as stated, will be required during the period '03 Q2 to '04 Q1:
| Task | Resource/ | Period | Deliverable |
| staff-years | deadline | ||
| SX/MPI software | 0.5 | '02 Q4 thru '03 Q1 | end Mar '03 |
| Install DBMS | 0.3 | '03 Q2 | end Jun '03 |
| Stacking/mosaicing/adv. pipeline | 0.5 | '03 Q2 thru '04 Q1 | end Mar '04 |
| Pairing/indexing | 0.1 | '03 Q2 thru '03 Q3 | end Sep '03 |
| Calibration SW | 0.1 | '03 Q4 thru '04 Q1 | end Mar '04 |
| Remote client SW & GUIs | 0.5 | '03 Q2 thru '04 Q1 | end Mar '04 |
| `V0' curation & interface tools | 0.3 | '03 Q3 | end Sep '03 |
| Total: | 2.3 | ||
The above resource estimates include allowance for coding, testing and documentation. They are approximate but should be close, based on our experience and existing knowledge of the size of the WFCAM science archive problem. Obviously, we will have a clearer idea of the requirements at the end of the analysis/documentation phase, i.e. by the end of '03 Q3. The next WFAU rolling grant review (mid-2004) provides a natural and timely review point for assessing completion of the analysis/documentation phase, and reassessment of the resource requirements in '04 to '06.
Critical milestones in this phase are WFCAM first light (end Q3 '03) at which point we intend to have a `Version 0' archive set up to allow users access to the database via a simple interactive interface, as data begin to be ingested. Then, at the end of Q2 '04 we intend to have our full-blown science archive ready for use. Included in the work to this milestone is the external project on astrostatistics and informatics (see Section 4b) Such issues as advanced machine learning classification techniques and spurious object detection are being developed, and we intend these to be applied to WFCAM data. Furthermore, we anticipate added value to the software programme from VO initiatives - for example in the area of datamining tools.