Monday, March 07, 2005

Terry Wallace, ADSR EmailGram

From Terry Wallace:

ADSR Emailgram
March 7, 2004

In the next month and a half all the technical Divisions in ADSR will
undergo their Division Review. This process is a very important
element of assessing the quality and health of science at LANL. The
DRC "grades" the Divisions, and that information is rolled up to the
University of California, and ultimately, to NNSA. The final report
card for LANL performance is based on Appendix F of the prime
contract between UC and NNSA, and the grades impact that award fee
given to UC. For the first time in history, UC received a
substantial fine based on their FY 2004 report card (sort of the
opposite of paying your children for every A they bring home). ADSR
owns the part of the report card grading science excellence. It was
one of the few areas in which LANL was graded as outstanding last
year. One thing is clear looking at the DRC materials for C, EES,
MST and T divisions: I expect that the performance will be
outstanding again. I don't believe that there is grade inflation in
this assessment. LANL has a very specific mission (National
Security), and constraints on how work is conducted. Despite these
obstacles, LANL continues to produce more peer reviewed publications
that our sister labs, and has numerous national awards given to its
scientists. We are at a significant juncture in time though - unless
we improve the environment for science soon we will begin to start a
downward spiral, losing key personnel and research focus.


Contact Me: I encourage you to email me with your questions,
comments, or concerns. My email address is, or if
you prefer, you can send email to my home address with confidence of

Terry C. Wallace, Jr
Associate Director for Strategic Research
Los Alamos National Laboratory

Phone: 505-667-8597
Fax: 505-667-5450

On balance this is a positive development. At least he admits that problems are coming. There are some caveats to consider.

(1) I am sure that many of us reading the blog would agree that there are systemic problems with “grade inflation” associated with DRCs. At the very least we’re being graded on a curve. It certainly been my experience that the DRC process leaves much to be desired especially with respect to conflict of interest. It would have been far more positive had Wallace commented that problems exist and are being addressed rather than defend a flawed process.
(2) Wallace makes the implicit assertion that our National security mission is an obstacle. This is troubling. Our mission should be benefit. Our mission should provide research with a focal point. We are not the NSF, we need to use the National security mission to our advantage. The stategic research part of the Lab must be strategic rather than just some quasi-academic playground. I do believe that many in ADSR make significant contributions to the mission, but not enough. Too many see the mission as something to pay lip service to.
(3) He then goes on to tout the aspects of our performance that comprise the “window dressing” aspects of ADSR. It is “the look how great we are because we’re on the cover of Nature,” therefore our science in support of National security is great, ruse. It is nothing more that bait and switch.
(4) Finally, the “positive part of the note” about our dim future. The only problem is that the dark future is upon us. The downward spiral is already here with all the ingredients that Wallace notes. It’s been with us for a while and Nanos has helped increase the rate of decline.

The time for action is now, it may already be too late.
The downward spiral of decreasing TSM personnel, scientific reputation, research funding, and morale began the day Nanos had a temper tantrum and shutdown the entire lab from a 'cell phone in his car' (to paraphrase the all powerful lab director).
The first comment addresses a serious problem with the DRCs, grade inflation and lack of independence. As Wallace notes the DRC reviews become fodder for the UC evaluation of science at LANL, and this in turn drives the annual evaluation. DOE does not have an independent scientific opinion, but they should. The DRCs are NOT appointed by UC, as they should be, but by LANL management. Wallace makes an inaccurate comparison. It really is like letting the kids grade themselves. Reform is needed, and the recent Sandia/UT arrangement looks useful; with UT making independent evaluatins of Sandia science programs.
Another point Wallace misleads us on is the published papers. LANL "buys" a lot of papers with LDRD money used for post docs, and much of this research is not "mission related". This money should come out of the PR side of the Lab. As created, the LDRD money was for "mission related" research, and the DOE should insist on this. Sandia, once again, looks much better on this. Openness is another issue with LDRD, as LANL, unlike other DOE labs, hides the LDRD program from the public. A corrupt "black hole" which consumes $100 million per year. Bowles promises reform, as others have in the past. The only true reform will come with openness.
Wallace seems well intentioned but does not acknowledge the very real problems, or propose solutions. Not a good start.
I want to comment on the impression that the mission is viewed as an "obstacle". There seems to be a divide here, between X and DX on the one hand and SR on the other. Other posters have often made comments that the basic science here is some kind of game, or waste of money.

This could easily degenerate into a flamewar, so let me just briefly state the prejudice that many in SR have towards X and DX, while acknowledging that it is a prejudice. This is the belief people in X and DX are not "real scientists". This belief is easy to understand: in many cases the degrees those TSMs have are from much worse schools and they have almost no record of published papers. Of course, they are not supposed to be publishing what they are doing, but even when talking to them they again are not able to divulge what they are doing and what they can divulge often seems simple. But this is of course a prejudice: the typical SR staffer has simply no way to judge the value of the work in X and DX.

Thus, the good thing is that the LDRD program is perhaps the easiest program for a person outside the lab to evaluate, contrary to the statement that it is a "black hole"! The work gets published and can really be checked.

However, there is a deeper problem behind this. The LDRD work should be more relevant to the mission. It was easy in the 1940s to get top scientists to work directly on the mission because the problem was defined. We've got to build something no one has ever built before and to do it we need to understand such-and-such, now, go do it. Nowadays, there is no clear direction on how the abilities of SR can be used. When I have tried to see how my research can be useful to specific programs I am often confronted with lots of delay, hesitation, and especially unwillingness to state a clear problem that I can solve! Some of this seems to me to be a desire to protect "turf". Whatever the reason, this is a problem!

In general, the real problem at the lab is a lack of a clear sense of what the mission is and how to evaluate whether something contributes to it. At a corporation this is easy: you make money or you don't. Here, the concern is that no one outside is evaluating the success of work in X and DX.

A colleague of mine, a basic research scientist, was asked to solve some problem as part of work on a program. The problem was solved and the response was: "well, we didn't want you to solve the problem, we wanted you to solve it with this code to check the code and you didn't do it that way so that's no good". That's just stupid. The original problem set then for my colleague should have been whatever problem the code was meant to solve. Or if that was classified, the general goal of getting working code in certain regimes should have been specified. These people missed a great chance: they had one of the few people at the lab or in the world who could have used experience in basic research to improve the code, and they wasted this resource.

What is missing is communication. In order to use SR effectively, clear problems must be stated. Sometimes, SR must allocate money to tasks that might seem like a waste of time in order to build capabilities to solve these problems. This is very necessary, because the kinds of problems that SR is best for are the problems that have never been solved from a fundamental standpoint, not the problems that are like engineering. Such problems can only be tackled from fundamentals. The staff in SR need to realize that a good scientist can find interesting basic problems almost everywhere, but those outside must realize that the only good use of a basic research scientist is on problems that he finds "interesting". Not because of some prima donna belief that "I can do whatever I want!", but because that is the unique ability that these people have!
I'm reading about how ADSR got an "Outstanding" for Science Excellence and expects to get the same this time around. LANL has more peer-reviewed publications than its sister labs, etc.

Then all of a sudden I see this big warning about being at a significant juncture about improving the environment for science. If there were deterioration in the environment, wouldn't one expect a less than "Outstanding" grade, or at least some sort of measurable impact?

In any event, sounds like the Divisions in ADSR are holding their own and doing the best they can. The fact that NNSA doesn't seem to value it much in the Award Fee to UC isn't their fault.

In fact, I would not be surprised in the future if an "Outstanding" in these areas will border on meaningless as far as the contractor fee goes.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?