Astromanager LLC

Our latest blog post:

Tag: Big Science

  • Why are we surprised when vendors can’t keep projects on time and budget?

    It has been occurring to me lately that there is a fundamental mismatch between how some astronomy projects  select and manage external vendors and their expectations of these external vendors’ performance.  In our case, we require competition and fixed price contracts.   Our typical instrument contract bidder, a university lab or department, is often not used to fixed price contracts.  To inappropriately stereotype, academia likes to be awarded grants and submit reports on the work accomplished after the fact.  That’s not what our contracts are. Universities are not used to, and therefore do not generally have the infrastructure for, fixed price, fixed schedule, competitive bid contracts.  They have not usually developed the considerable expertise needed to fully cost a new project and identify and manage  its  risks.  The end result is that they often deliver late and and over-budget, if at all.

    An alternative to academia. we could pursue our contracts in the commercial world.  Except for  one, they don’t usually respond to our requests for proposals, and two, they generally cost more.  Their more properly calculated and risked cost is what generally prices them out of the competition and prompts them not to waste their time bidding.  Again, to inappropriately stereotype, the commercial world has learned how to cost and scope a project to ultimately complete a risky project to a fixed deadline and cost.   While their bids end up more expensive, the price more accurately reflects the risks involved in a fixed price bid for what is ultimately, after all,  a one-off product.  Those that know how to properly cost an uncertain project, whether in academia or outside, will inevitably come up with a higher price and a later delivery that those that do not.  So we are simply fooling ourselves if we expect a complete, on-time, on-budget delivery after selecting the low bid from an ill-prepared institution.

    So, we have a couple choices. We can pay the higher price requested by those that really know how to cost a risky project or we can adjust our approach and our expectations and work with less-prepared institutions.  The latter, though, means acknowledging early on that the bid cost and schedule will not likely last the whole project.  If our selected vendors do not have the tools for full project management, we need to work more closely with them throughout the project, both training them and working as embedded project managers.  This effort will have real costs in terms of both resources and schedule.  By using our contingincies (see some of my thoughts on the use of contingency in the middle of this earlier post)  properly, for example,we can begin to generate a culture of cost and schedule containment from the early stages of the project.  At the beginning of a project, when the first hurdle is found, most people want to throw time and money at it since at that point, these reserves are well-stocked and it seems too early in the project to de-scope.  Yet, if schedule and cost are to be maintained, de-scoping early is often the most appropriate response.  One way to soften the blow is to leave hooks for the de-scoped capability to be added again later, if reserves allow. In some instances, however, that simply won’t be possible and a given capability will simply have to be eliminated in order for the project to stay on track.  Making these kinds of decisions early will help the project develop a plan-oriented culture.  If, on the other hand, there is more time and money than there is functional contingency, then we need to realize the initial optimistic delivery dates and costs  are not likely to actually correspond to the final delivery dates and costs.  We therefore need to set our expectations, and those of our community, accordingly, by adding in the schedule and cost contingencies necessary to assure the required functional specifications are met.  If time and schedule reserves are not very plentiful, then we must prepare our stakeholders for an ultimately de-scoped end-product.

    If we really do need something on a fixed price and schedule, then we need to consider paying more to the people who know how to deliver in those circumstances.   It will look like it costs more to do so, but in the end, it may actually cost less and is certainly more likely to result in a useful delivered product than would a lesser contract awarded to a less realistic vendor.

    (I realize there are commercial companies who don’t understand risk and universities that do, but my point here is to understand the vendor and adjust your stakeholders’ expectations and project management approach, accordingly. We may not always be able to choose with whom we work, but we can choose how we work with them.)

    Scot has managed both projects with both commercial and university vendors. He’s seen similar mistakes made by each type (not reading the contract is a common one), but has also learned to adjust his style dependent on the vendor’s experience and approach.

  • The never-ending battle between big and small science in astronomy

    When you hear the thundering herd behind you, it’s time to move on to a new field. That’s advice I often heard from my graduate advisers. While you might call this a “small science” mindset, they went on to found the Whole Earth Telescope (WET), an international collaboration of astronomers at more than a dozen observatories around the world that coordinated observations to follow variable (white dwarf) stars continuously for up to two weeks at a time, clearly a “big science” approach. The WET was initially very successful, and began to falter only later as it struggled to transition from a bunch of astronomers doing what they needed to do to get their science addressed, to an institution looking to continually justify its funding and purpose.

    I recently finished Giant Telescopes by Patrick McCray, a book basically about the origins of the Gemini Observatory. I was struck at how many of the same arguments that were in the community decades before Gemini, persisted up to and through its construction and are still being debated today. Principally, the question of private vs. public and big science vs. little science. In my earlier posting about The role and need for an international observatory, I gave some of my thoughts on the first question so here, I want to at least introduce the latter.

    A few years back, Simon White and Rocky Kolb submitted a set of papers, each championing for the big science or little science models for astronomy. There was even a pseudo-debate between them at the 2008 AAS meeting (http://aas.org/taxonomy/term/27 – session 87 – where you can see a video of the discussion). Simon White’s paper was Fundamentalist physics: why Dark Energy is bad for Astronomy while Rocky Kolb’s, issued in response, was entitled A Thousand Invisible Cords Binding Astronomy and High-Energy Physics. The context for this particular discussion was Dark Energy, but the underlying issue was really whether or not astronomy should be done in a big science or little science approach.

    An artistic interpretation of the crystallized white dwarf star, BPM 37093, observed by the WET.BPM 37093 is so massive, that theory predicts its core, mostly carbon and oxygen, is crystallized. Here on Earth, crystallized carbon is called diamond. Observations of the oscillations of this star with the Whole Earth Telescope were consistent with this interpretation and placed strong limits on the amount of crystallization within the star, a diamond in the sky.

    I don’t think this argument will ever really die since we will always have competing projects that are each done best under a different model. The solution is going to be to continue to adapt and be aware of the compromises and needs necessary to keep both approaches viable. One interesting moment in the 2008 AAS “debate” was when an audience member asked what each would like to adopt from the other side. Simon White said of high energy physics “managing large projects” while Rocky Kolb said of astronomy “making data public”. What I liked about this question and its responses was that it acknowledged that we don’t have to simply emulate the high energy physics big science model, nor steadfastly stick to astronomy’s traditional small science mode, but we can learn from both and make something better than either alone. The Sloan Digital Sky Survey (SDSS), like the WET, is a good example of this kind of approach. A core group of people inspired and really implemented the survey, with formal management and technical support partially adopted from the particle physics world. The SDSS used both public and private funding and made all the data publicly available after a short proprietary period. This melding of approaches helped make the SDSS one of the most successful projects of its type and certainly helped pave the way for even larger projects like the LSST and PanSTARRS.

    So the question isn’t big science vs. small science in astronomy, but how do we create an environment where both can exist, cooperate, and thrive? With 30m telescopes, 8m surveys, and pushes to build large, wide-field survey imagers and spectrographs, astronomy must learn to embrace big science, although we can do so on our own terms, not necessarily on those laid before us by other fields and previous projects. This debate is similar to the one on public vs. private facilities. A true strength of the astronomy community is that both public and private facilities have been successful. That both are continuing to debate why they each need more resources than the other means the community is relatively healthy. The next hurdle in both these arenas will be how to ensure the appropriate levels of cooperation between each community. How do you motivate private funding when the data become public to all? How do you (or do you?) justify public funding when the resulting data remain private? How do you make sure individual contributions are visible and not an anonymous contribution to a juggernaut project? How do you handle risk in a extremely delicate, risk-adverse, funding environment, especially in a field which traditionally pushes at the outer limits of available technology, a fundamentally risky task?

    I’ll try to address some possible answers to these questions in future posts.

    Starting off with the Whole Earth Telescope, then onto the SDSS, Subaru, and now Gemini, Scot has been involved in increasingly “larger” science, but has always managed to come away with his own “small” science projects within each. He particularly enjoyed doubling the number of known white dwarf stars from SDSS data of largely failed attempts to find Quasars!

Big Science Boards Books Budgets Business Change Management Classes Communications Community Competition Conflict Crucibles Emotions Employee Development fire-fighting Fun Gemini GTD Hiring Impostors International Introduction Leadership Management Management Corollary Matrix Mauna Kea Federation Mistakes Politics Productivity Project Management Scientist Dilemma SDSS Social Media Videos Whole Earth Telescope