Project Manager (Owner)
IS Applications Division
IS Applications Division
Production Management Coordinator
IS Applications Division
Business Area Manager
Governance and Strategic Planning
was Nigel Brown
|Lead Developer||IS Applications Division||Alyson Shaw|
|Business Area Manager||Governance and Strategic Planning||James Toon|
IS Applications Division
|Technical Architect||IS Applications Division||Mark Lang|
The main aims of the project were to: • replace the Publications Repository and the Edinburgh Research Management Information System (ERMIS), combining and enhancing their functionality • enable administrative oversight of research activity data to ensure its accuracy and therefore the accuracy of statutory returns to bodies such as SFC and HESA, and to exercises such as the Research Excellence Framework (REF) • strengthen the University’s research management processes by enabling the incorporation of research activity data • enable the collation of data with research collaboration partners (e.g. pooling), facilitating management and preparations for external research assessment • give researchers access to their research activity data for use in their CVs and web profiles • improve the University’s communication of its research activity, and hence widen its impact
The key driver for the project is ensuring that we accurately capture the data required to make our REF2014 submission in 2013.
At the closure review meeting it was agreed that the project delivered what it set out to do. Despite some elements of the project being delivered later than scheduled, it was agreed that the project was a success, with GaSP, IS Applications, Library, Finance and HR all contributing to the success of the project. The closure review also focused on an lessons learned from the project. The following points were noted:
What went well:
- The dedicated project team assigned to the project during the first phase worked well. - The relationship with the software vendor Atira (both technical and business) was good and helped contribute to meeting the deliverables. - The iterative approach to developments was positive. The business areas could test components early as opposed to traditional approaches to software projects. - Project governance was noted as being successful. The Steering Group was engaged and made decisions when needed. - Risk management. The Steering Group took ownership of risks when necessary. GaSP are using the risk management techniques for future projects. - Positive feedback has been received from Schools during the roll-out of the service. What could be improved: - The project could have been split into two projects. 1 Software installation and publications service. 2 Finance/VRS extracts. (Note, Project Services will be making recommendations regarding Project Phases, Sub Projects and Work Packages). - Problems were encountered when the first phase of the project went live and into service. Support vs Project caused some issues (e.g. production and project needing to work on the same database packages). - VRS extract. As noted below some scenarios were difficult to handle. GaSP and IS will review the extract to see if any improvements could be made now the new IDM is in place. - Technical environments. PURE currently has BETA, TEST and LIVE environments (BETA replacing DEV). GaSP have enquired about a pre-production environment. IS Applications will be reviewing this across other university systems.
Objectives / Deliverables
The objectives and deliverables stated in the project Terms of Reference (ToR) are shown in the table below:
Notes from Closure Meeting O1 - D3: as agreed with the Steering Group the data feed from EUCLID was not included in the scope of the project, and was deferred to a future project. - D6: it was noted that the publications data migration tasks (A) took more effort from the Library than originally estimated. - D6, B iii: The link to the university's Open Access repository (ERA) was removed from the scope of the project, as agreed with the Steering Group. O2 - D1: The research financial info from eFin was split into two deliverables (R code extract, and Non R code extract) - D3: Extract from VRS. It was noted that it was difficult to manage scenarios where researchers had a combination of HR and VRS ids. A number of manual steps are required in order to show the correct information in PURE.
To deliver a Publications Service by 31 July 2011
Procurement, installation and configuration of PURE system hardware.
IS Apps and IS ITI have agreed the hardware requirements for running PURE. Additional hardware for the production environment has been purchased according to Atira’s recommendations.
The initial developments will take place on Atira’s infrastructure in Denmark (data protection agreement signed by Atira). In mid June, following successful testing of HR/Org data and publications data import from UoE repository, the PURE software will be installed on UoE infrastructure. Atira will assist IS Apps with the installation on one environment. It will be IS Apps responsibility to install PURE on the remaining environments.
Creation and configuration of a data holding area for integration with PURE. Data will be extracted from the following sources:
- Oracle HR
- Organisational Hierarchy
- Visitor Registration System (VRS)
This data is required for research management purposes and is key to the REF submission.
Import of the following items of data from existing sources:
i. Date of entry into the academic profession [static data from HR]
ii. Research Assistant | PI Associations [static data from HR]
iii. Research Press Coverage [MS Access Database]
Note that following import of this data, PURE will become the golden copy of the data from then on.
Identify user roles (e.g. researcher, research administrator, cataloguer, etc.) for use in PURE.
An individual’s role will define their interaction with PURE in terms of the data available for viewing, updating, etc. Users can have more than one role (e.g. researcher, Head of School). Proxy roles will be available.
Workflows for different data types (e.g. publications, updating research grant data) will also be defined and implemented.
Deliver publications functionality
A – Publications Data Migration Tasks
i. Extract of publications data from the existing Publications Repository, the Edinburgh Research Archive (ERA), and other local publications data acquired through direct liaising with Schools.
ii. Data cleansing, transformation, and upload of publications records into the PURE .xls templates, in preparation for legacy import into PURE.
B- Publications System Functionality
iii. Link PURE to the University’s Open Access repository ERA and reorganise the ERA collection structures to accommodate data transfer from PURE.
iv. Transfer of metadata only records held in PURE into ERA for public access, in addition to full text records (still to be determined whether this will happen in Phase 2 or 3).
v. Once PURE goes LIVE, users will be able to scan citation databases (TR Web of Science, ArXiv, PubMed, and WorldCat) to harvest bibliographic metadata directly into PURE, thereby eliminating the need to re-input any data. The list of external citation databases available for scanning will expand as more UK PURE customers add connectors.
vi. Develop processes and workflows for deposit of metadata and full‐text files, including documenting cataloguing requirements and developing a strategy for reviewing and validating new items.
Once PURE is operational it will become the golden copy source for Publications Data. The in-house Publications Repository will then be decommissioned.
Researcher Web Profiles available
With HR and publications data available in PURE, basic online researcher profiles will be available via the PURE Portal. The PURE Portal allows external access to selected PURE content. Users will be able to choose which content they wish to make publicly visible through the Portal and a roll out process will be agreed to ensure sufficient population of the Portal.
Researcher CV functionality available
With HR and publications data available in PURE, researchers can use the CV module to build personal CVs.
Roll Out Strategy
In parallel with Phase 2 of the Implementation Project, the Project Team will define a Roll Out strategy. This will include:
• Training and communication strategy (e.g. School‐by‐School roll out, or identify local ‘super users’ to train first, etc.)
• User engagement expectations (direct access or proxy users)
• Agreeing incentive schemes to encourage deposit of research activity data (e.g. appearance on UoE homepage).
Service model and support arrangements
The roll out of the PURE system will require new roles to be defined for support purposes. The Project Team will define the support arrangements and service model.
To deliver Research Management Information by November 2011
Research financial information from eFinancials
Research grant details (i.e. job code, PI name(s), PI split(s), source of funding, annual expenditure, award amount, etc.) back to 1 August 2001. Research grants will be linked to individual researchers using the PI HR/VRS ID(s) and PI split(s) in eFinancials.
Data from eFinancials will be loaded to PURE on a nightly basis. eFinancials will remain the golden copy source of research financial data. Any changes to financial data held in PURE can only be done via eFinancials.
ResearchBuildingProject Income Spreadsheet
GaSP, in liaison with Corporate Finance, maintains a spreadsheet detailing research building project income back to 1 August 2001. This is updated annually, and will therefore feed into PURE on an annual basis.
Researchactive visitor details (e.g. name, visit dates, organisation, etc.) for visits in effect after 1 August 2001. Using RAE2008 UoA assignments (last data quality check was undertaken in November 2007), PURE will identify an initial population of research‐active visitors. Users will be able to enrich this list by identifying any additional research active visitors.
Data from VRS will be loaded to PURE on a nightly basis. VRS will remain the golden copy source of research active visitor data. Any changes to research active visitor data held in PURE can only be done via VRS.
Capture esteem indicators, events, and other knowledge exchange activity
PURE will become the ‘golden copy’ of the following data items, to be input by Schools / individual researchers:
• Esteem indicators (prizes, awards, invited lectures, journal editorships, board memberships, etc.);
• Events and other knowledge exchange (KE) activities.
Aggregate reporting available
From 31 October, research managers and administrators will be able to use PURE to report on data items at the aggregate level (i.e. School, College, University) back to 2001/02.
Data clean‐up reporting available
From 31 October, research administrators will be able to use PURE to report on data for clean‐up purposes (e.g. research grants with no PI details input, etc.).
Researcher Web Profiles available (Phase 3)
With the full data set available in PURE, on‐line researcher profiles will automatically include the new (Phase 3) streams of data.
Researcher CV functionality available (Phase 3)
With the full data set available in PURE, researchers can use the CV module more fully to build comprehensive personal CVs.
The scope of the project as defined in the ToR was:
The project has two main objectives:
1. A Publications service by end of July 2011 which will be met by the deliverables D1 – D9 (O1).
2. Research management information by end of November 2011 which will be met by the deliverables D1 – D8 (O2).
Clarification 1 - Student data:
The extract of student data from EUCLID to PURE is not in scope.
The current PURE UK data model for student data is a simple one. Atira have plans to revise the model in light of the REF2014 requirements, but no further work is being done in this area until these requirements are known (July 2011). The PURE Project Team presented two options to the Steering Group regarding progressing with the student data extract:
1. Delay the extract of student data into PURE until the UK student data model is updated to capture more full information from the student record (as per UoE functional requirements).
2. Extract student data into PURE in this initial stage, with a view to re‐writing the extract as per an expanded UK student data model as soon as Registry / IS resource is available (2012/13).
The Steering Group recommended that option 1 was adopted and this project will follow the recommendation.
Clarification 2 - Decommissioning of in-house systems:
The objectives state that PURE will replace the two in-house systems – the Publications Repository and ERMIS. The project deliverables will allow these systems to be decommissioned, however the actual decommissioning of the systems is not in the scope of this project.
Notes from Closure Meeting
It was noted that the project stayed within the original scope, with some deliverables being removed as described in Objectives section above.
The original schedule as defined in the ToR was:
The implementation of PURE at the UoE will involve development work at Atira’s site in Denmark as well as at the UoE. This can be summarised as follows:
1. A data holding area will be created on UoE infrastructure to store UoE HR and Organisational data (completed during RES035 project).
2. Atira will create a development version of PURE on their infrastructure. They will remotely connect to the data holding area at the UoE and import the HR and Org data (completed during RES035 project).
3. An export of Publications data will be provided to Atira which will be imported to their development system. This data will be integrated with the HR and Org data.
4. Once the initial development on Atira infrastructure has been completed, PURE will then be installed on the UoE infrastructure.
5. Following successful testing on UoE infrastructure, the PURE publications service will go live at the end of July 2011.
Type (& Reason if Required)
HR Extract Analysis and Design Review
Software Installation Acceptance Review
HR Extract Acceptance Review
PURE Publications Functionality Acceptance Review
Deployment to Live
Research Management Information
Finance Extract Analysis and Design Review
VRS Extract Analysis and Design Review
Finance Extract Acceptance Review
VRS Extract Acceptance Review
PURE Reporting Acceptance Review
Notes from Closure Meeting
- The first phase of the project stayed within schedule and was delivered in the summer 2011. - The second phase of the project was not delivered according to the original schedule. The Finance (R Code) and VRS extracts were delivered in Feb 2012. The Non R code extract was delivered in Feb 2013.
Analysis of Resource Usage:
Staff Usage Estimate: 441 days
Staff Usage Actual: 677 days
Staff Usage Variance: 54%
Other Resource Estimate: days
Other Resource Actual: days
Other Resource Variance: 0%
Explanation for variance:
The project estimate remained on track during the first phase of the project, which was delivered on time and within budget.
Additional effort was required to complete the finance extracts (both R code and Non R Code). The non-R codes were particularly problematic since there was no clear specification . They were less standardised than the R codes and it was more difficult to work out the definitions for the various pieces of data. . Changes in developer staff during the project also resulted in additional effort being required. This was needed for handover and gaining knowledge of PURE, as well as the finance extracts.
The effort breakdown by team is as follows:
Project Services: 139 days Config Team: 383 days Dev Team: 3 days Dev Tech: 126 days Service Man: 14 days Production Man: 12 days
Key Learning Points:
1. Some low priority JIRA calls (relating to data extracts) will be delivered by the RES039 PURE Enhancements project. 2. The PURE Upgrade project (RES048) will make further changes to the technical einvironments, as well as a performance review of the current Pure service.