Gemini Lessons Learned Workshop

The Gemini Dedication took place on June 25 after eight years of effort from an international collaboration. Canada is a partner in what is arguably the world's best telescope but there will be only one 'facility' instrument when the first call for proposals is issued later this year. The Keck telescopes have experienced similar lost scientific opportunities because of delayed instruments and instruments that were built but never reached the telescope.

Why are instruments for these new, large telescopes being delivered late and over budget? What can we learn from our experiences that could help us to avoid repeating the costly mistakes of the past?

On behalf of the Gemini partnership, Andy Woodsworth and Tyson Hare of the Canadian Gemini Office organized a workshop in Parksville, British Columbia on July 8 and 9 to explore these issues. The 31 participants included instrument builders and others from the Gemini partners, UKIRT, Keck, University of Florida, AAO and UH, were very frank and open in discussing their experiences with building instruments for large telescopes. The full proceedings have been prepared by Tyson and Andy, and will likely become available on the Gemini Web site in the near future.

The common themes that emerged from the meeting are succinctly summarized in the proceedings as:

• The importance of good project management
• The need to learn and use "design to cost" techniques.
• The importance of trust, respect and good communications in partnerships
• The need to provide sufficient incentives for groups to want to build instruments for Gemini, given the potential for severe financial problem
• The need for rapid conclusion of contract negotiations
• Difficulties with the EPICS software causing schedule overruns

To expand slightly on the first two points:

'Design to Cost' Issues

There is a need to balance science and performance capabilities, innovation and creation, and cost and risk. Detailed engineering needs to be considered with science requirements from the beginning planning stages. 'Closed loop' design and scheduling is required. Small changes in scientific requirements, workscope or schedule can have a significant overall effect on the project cost and deliverables.

Project Management Issues

The participants concluded that effective project management is essential to the success of a large ($3-6M US) instrumentation project. There was much discussion over what constitutes 'real' project management. It was concluded that good project management is able to:

• review effort and catch problems before they become too serious
• track progress and make sure targets are in sight
• track budget, schedule, objectives, goals, and requirements
• look at the 'big picture' in a continuous manner to:
• distribute the workload in an effective manner
• determine the effects of changes to the delivery date
• determine the effects of trade-offs
• determine the critical path and the 'reasons' that a path is the critical one

Everyone that attended felt the Workshop was very beneficial and Gemini is discussing the idea of holding similar meetings in the future.

This meeting was Andy's last major contribution to the Gemini project. The Canadian Gemini Office extends its warmest thanks to Andy for his very capable management of Canada's participation in Gemini.