Experimental User Survey

Total Entries = 85 Status Grad Student = 6/85 Post-doc = 10/85 Professor = 39/85 Research Staff(Non Jlab) = 15/85 Jlab Staff = 8/85 Other = 7/85 Primary Hall Affiliation Hall A = 35/85 Hall B = 24/85 Hall C = 18/85 1. Proposal Process Question Strongly Agree Agree Neutral Disagree Strongly Disagree a The requested documentation to acommpany my proposal is simple to complete. 4/85 30/85 20/85 6/85 0/85 b Submitting my proposal to the user liaison office was easy. 12/85 39/85 9/85 1/85 0/85 c I found the Technical Advisory Committee comments to be on the mark. 6/85 29/85 19/85 4/85 1/85 d I found the Theoretical Advisory Committee comments to be useful. 3/85 23/85 24/85 5/85 2/85 Documentation for the beam time requests was not very clear, and sometimes different versions were inconsistent (email that was sent vs. info on web page vs. info on the forms). Additional guidence for breaking up the run into different settings (for the scheduling and for radiation budget) would have been useful. Didn´t receive Theoretical Advisory Committee comments. Wasn´t even aware that this committee existed Neither committee had anything useful to say; it would be best to do away with both committees. The Tech and Theory Committee comments apparently are each comments from one individual. That individual should be identified. The Theory comments are quite subjective, and it would be useful to have comments from a real committee of say three. n/a I had no experience with the technical advisory committee. I have never heard of the theoretical advisory committee. My proposal was in 1994, maybe that´s why. I leave this section to the experiment spokesman I do have strong reservations about the flavor of the subjective component entering the grading of experiments. I do not always agree with the commnets from the TAC or other committees, but they are useful. never submitted proposal I note that one might find comments to be useful without agreeing with them. In my experience the quality of c and d varies quite a bit View Comments 2. The PAC Process Question Strongly Agree Agree Neutral Disagree Strongly Disagree a I found the PAC members to be qualified to judge my proposal. 7/85 35/85 12/85 4/85 1/85 b The PAC made every effort to try to understand my point of view on issues related to my proposal. 4/85 35/85 12/85 6/85 0/85 c The PAC's questions about the proposal were reasonable. 5/85 37/85 13/85 2/85 0/85 d It would be desirable to have written comments on the proposal from the PAC one week before the PAC presentation. 23/85 19/85 12/85 2/85 1/85 Why does it take months for the PAC report to become public? There should be time limit of a week. Have not personally submitted a proposal for some years This is asking a lot of the PAC Early comments from the readers would obviously help in preparing an appropriate oral defense. Many PACS have written comments as standard practice. Why not JLab? They can lead to more sharply focused presentations and thus to a better discussion of the real issues. n/a I think it would be very beneficial to have a written explanation of why the PAC assigned a given scientific rating to a proposal. The ratings are at the heart of the activity and have an important effect on scheduling. I leave this section to the experiment spokesman I have thought for a long time that the PAC was biased towards particle physics and away from everything else (the well-known 1/A effect). An effort should be made to provide more balance on this important committee. The PAC seems to support physics that was done already at places like SLAC, Bates, Mainz, NiKHEF, etc (just reducing error bars or extending the measurement a bit in kinematic variables) over supporting new and novel ideas never tried before. There is a need for a bit more balance between ´bread and butter´ physics and the exotic new stuff. idem The PAC decisions look very incoherent, from a physical point of view While I don´t believe written comments from the PAC before the meeting are needed, it would be nice if the PAC would finish writing their final comments while they are still togehter, instead of taking months to complete the document via e-mail. If there is a question 2d, it does not show up in my browser. I thought we are taught that all questions are reasonable, but not all answers. Also, I find that the PACs have tended to be *slightly* incapable of judging some experimental feasibility issues, though overall better than myself, of course. Don´t PAC members contact the spokesperson(s) if they have questions that should be looked into? View Comments 3. Scheduling Question Strongly Agree Agree Neutral Disagree Strongly Disagree a The process of scheduling experiments at Jefferson Lab is clear and open. 6/85 23/85 22/85 17/85 4/85 b The flexibility built into the present schedule, ie, the 'float',provides ample beam days to complete the approved PAC days. 6/85 30/85 22/85 12/85 2/85 c The jeopardy process is needed to ensure that the best experiments are run in a timely manner. 17/85 33/85 15/85 6/85 1/85 d I think that 3 years is the optimum jeopardy cycle for the Hall in which I do most of my research. 6/85 20/85 28/85 17/85 2/85 I think that the way jeopardy has been handled has caused siginficant problems. However, this is not due to the process in general, but the fact that many years worth of jeopardy proposals were brought back in one year. The only thing that this seems to have accomplished was a reduction of the beam time committed on paper, and the requirement that experiments that were approved, and were still good enough to be accepted had to or will have to come back multiple times because of PAC requirements to trim back the beam time, independent of quality of the proposals. Under normal circumstances, 3 years is probably a reasonable time frame. experiments which require considerable preparation and are not even intended to be scheduled for 2 or 2 years should not be subject to jeopardy 3 years is too short. The lab cannot deliver either the advertised current nor the energy; this impacts the physics intensely. Before "upgrading" from 6 GeV we should first REACH 6 GeV and do physics. We do NOT have physics-quality beam -- we have "stunt" beam for one-time only photos. CEBAF could learn from Bates. Considering the time it takes to set up and run the average experiment, at least in HAll B, 4 years might be more reasonable. It should be longer. Sometimes it can take that long to get scheduled. Hall B experiments are of a different nature than A or C, and the jeopardy process does not appear to work well there. n/a To have a simple system, 3 years is probably ok for jeopardy. It´s not the scientifically optimal way, in that some experiments are in areas where there will be no new developments until the experiment is run, and the scientific interest has been a constant for 5-10 years, while in other cases a much shorter time scale might be appropriate. The PAC could choose a jeopardy interval which varies with experiment, or could have the option of choosing a number different from 3 years if desired. A little longer jeopardy period would be useful. The jeopardy time, if at all, needs to be lengthen to at least 5 years. Jeopardy cycle should be extended to 4 or 5 years. in era of declining graduate students and postdocs it is imperative that lab considers as scheduling constraints any preferences arising from university obligations of faculty spokespersons/pis I do not understand question 3a/ Scheduling of an experiment is to a large extent independent of its rating, and more contingent on the coincidence of favorable circumstances. Therefore the three year jeopardy cycle is too short in many cases. There are hardly any physics issues that become resolved in 3 years, and the users´s and PAC´s efforts involved in the frequent re-review process could find better use. Scheduling of experiments needs to be improved. Jeopardy should be 4 years I think a 4 year jeopardy time would be better. More flexibility is needed to deal with periods when the accelerator and hall equipment are down more than half the time. View Comments 4. Experiment Set Up Question Strongly Agree Agree Neutral Disagree Strongly Disagree a Jefferson Lab provides sufficient technical support for setting up experiments. 13/85 46/85 10/85 7/85 0/85 b Adequate time is allowed for experimental setup. 9/85 42/85 10/85 11/85 3/85 c The user community provides the apropriate support for setting up the experiment. 6/85 38/85 23/85 6/85 1/85 d Information on paperwork (COO, ESAD, and RSAD) needed before an experiment can run is easily accessible. 11/85 37/85 23/85 4/85 0/85 Setup schedules have generally be adequate, though not sufficient in case of significant problems. JLab and the user community have generally provided adequate support for any major setup tasks, but have not always been fully prepared to take data (trigger setup, data analysis, etc...) Haven´t set it up yet, so no comments n/a The user community connected to Hall B does not provide sufficient manpower for setting up experiments. As a graduate student I found the support structure very lacking. There is definitely a need for more hands on people from both the user and lab side. Unfortunately it is more difficult for a user to provide support since that support is dependant on the flow and number of interested graduate students. Other than the Hall C GeN fiasco, running of experiments is fine. the coo, esad and rsad are by most experiments just copied from previous ones, because otherwise it is a completely unreasonable amount of work. given that it is just a copy, it would seem that the exercise has more to do with behind-protection than anything else. All the setup periods end up being tight, it seems. I guss appropriately. Hall C leader is very supportive of providing resources for experiments. View Comments 5. Collaboration Training Question Strongly Agree Agree Neutral Disagree Strongly Disagree a The radiation training is useful for working on an experiment. 5/85 38/85 20/85 15/85 1/85 b The Hall walk-through is useful for working on an experiment. 11/85 49/85 13/85 4/85 1/85 c Have you taken MCC Operations Familiarization Training? Yes = 4/85 , No = 72/85 d (If you have taken the MCC OFT) I found the MCC OFT useful. 0/85 2/85 2/85 1/85 0/85 e The Conduct of Operations for my hall adequately explains the expected duties of the physics division liaison, the run coordinator and the shift leader. 10/85 51/85 14/85 0/85 0/85 f For the first-time experimenter, there is sufficient guidance on how to perform duties on shift. 1/85 38/85 20/85 16/85 1/85 g Sufficient training is provided for operation of cryogenic targets. 7/85 30/85 22/85 10/85 3/85 Documentation for novice users varies by hall and by experiment, and is scattered in many places. A set of basic how-to instructions, compiled in one place and kept up to date, would significantly help. Rad training is not immediately on-the-floor useful most of the time (which is good), but it provides important information to an experimenter. I think the form in which it exists is fine. The idea of a hall safety walkthrough is a good one. However, the current implementation is a complete waste of person-power. The walk-through should consist of a short briefing on hazards in the hall, and familiarization with what alarms are, what they all mean, and where the exits are. Each person should then be responsible for knowing what safety equipment exists in their working area. But the philosophy that every person needs to know what every single fire hazard is in the hall is a complete waste of time. If a safety briefing can´t be expressed in one side of one page, then it´s too long. In an emergency, you are only going to remember a few pieces of information, and that´s how to get out safely. The COO and associated documents border on having too much irrelevant ´boilerplate.´ They are not irrelevant documents, but neither are they very useful. The coo, etc is not a good place to learn for first time. a training shift is more appropriate. cryo training should give more emphasis on unusual situations. I would comment that there is much the shift people need to know, and it appears few really understand it from all the paperwork alone. And much of the details of what one does is not in this paperwork. 5f A first time experimenter should never be place in a position where he needs to rely on previous and written guidance. They must be scheduled with appropriate experienced personnel. An experment cannot be made a turn-key operation. View Comments 6. Experiment Running Question Strongly Agree Agree Neutral Disagree Strongly Disagree a Jefferson Lab provides sufficient support for running experiments. 9/85 51/85 11/85 6/85 1/85 b The user community always provides the appropriate manpower to run an experiment. 7/85 40/85 21/85 7/85 1/85 c The communication between your Hall and MCC is clear and allows the experiment to run efficiently. 11/85 50/85 10/85 7/85 0/85 d The communication between your Hall and the other Halls needs to be improved. 7/85 19/85 37/85 12/85 0/85 e The system of having a Run Coordinator for each Hall and a Program Deputy for the Accelerator is an efficient means of coordinating experiments. 20/85 52/85 6/85 0/85 0/85 f Computer facilities in the Halls are sufficient for online analysis of the experiment. 7/85 40/85 19/85 5/85 4/85 g The counting house facilities are conducive to efficiently running an experiment. 5/85 51/85 12/85 8/85 0/85 h The Hall staff is responsive to the changing conditions of the experiment. 14/85 45/85 11/85 4/85 0/85 There should be free coffee in the counting house. Procedures for a run, expecially emergency procedures, should be in a centralized place in the control room. It´s all well and good to have documentation on the web, but for some things, that´s not appropriate. In every control room, there should be a red three ring binder labelled "what to do when alarms go off", and ONE PAGE for each type of alarm inside each page. During our last run, we had VESDA alarms go off in Hall B. There are procedures for what to do if there is an alarm, and after digging around on the web for 15 minutes, I was able to find them. The procedures for that kind of thing are kind of nebulous. For example, we had a problem with the vaccum system in Hall B the last couple of days of the run. We didn´t know who to call, so we called the wrong person. There is, apparently, an established procedure for "engineer on call", but we (the experimenters) were never told about that specific pager, or what its number was. MCC does things without checking with the experiments and does not do what is promised, especially regarding accelerator work. A running experiment should have priority over accelerator studies. The pd doesn´t seem to get enough support/respect from mcc. Hall B mostly is incapable of any online analysis. A small inexpensive compute farm (less than $100k) could remidy this, but neither the Hall nor the CC is inclined to go with inexpensive compute solutions. See below. The manning of shifts is well coordinated in Hall B. An inappropriate burden of responding to problems is undertaken by the users. My comments reflect experience in Hall B ONLY! I would just love it if somehow there could be a little more space in Hall-B counting house. Room for a little kitchen upstairs might also be nice. Again, other than the Hall C GeN fiasco, the support is fine. Some Hall staff member show reluctance to come in nights and weekends in the event of equipment breakdowns. View Comments 7. Experiment Analysis Question Strongly Agree Agree Neutral Disagree Strongly Disagree a Assistance from JLab in getting access to my data was adequate. 7/85 43/85 17/85 5/85 2/85 b I expect Jefferson Lab to provide the following percentage of the computing power for off line analysis of my experiments. 100% 7/85 75% 43/85 50% 17/85 25% 5/85 0% 2/85 c The computing facilities at Jefferson Lab are sufficient to satisfy these expectations. 1/85 38/85 16/85 14/85 4/85 d The Hall staff provide needed support in analyzing the experiment's data. 5/85 34/85 20/85 6/85 5/85 The usage of the tape silo and batch farm changes fairly frequently, in order to keep up with usage demands. Better communication between the computer center and users would allow much more efficient use of the computer resources. If the computer center (or someone in the halls) maintained a set of basic scripts to retrieve and analyze a set of runs that makes optimal use of the silo, it would make it much easier to use the analysis farm in an efficient manner. Bandwidth to off site is unacceptably slow, 1.4 GB/hr is the highest I´ve achieved, I routinely to 20 times that at my home institution. Nobody on the Hall A staff maintains ESPACE; the best is that Ole archives releases and Nilanga helps optimize a few things (TOF, optics). Hall B has seriously insufficient computer resources. The whole model of data storage is painfully inadequate. The tape silo is slow, unresponsive and error prone. There is not enough disk space by factors of 10 to 100. There is barely enough CPU resources, but since data comes over insufficient network bandwith, the CPU´s can keep up, since the problem is in getting the data there. Software for Hall B offline analysis is an unprofessional mess of unreadable code and poorly functioning modules that give dubious results (the CLAS field is not even known properly.) When one group leaves there can be a hole in the needed experitze. The staff have to have experts on all phases in order to assume continuity. n/a The user community associated with Hall B does not provide sufficient manpower for carrying out needed tasks related to the analysis such as: calibration of detectors, studies to understand magnetic field characteristics, detector acceptance issues, etc. More disk space is needed. View Comments 8. What single thing could you suggest which would improve the process of doing experiments at JLab? Get the PAC reports out very quickly See above. The single most frustrating thing about working at Jefferson Lab is that the lab assumes that any experiment has many people on site dedicated to it´s running. This is not true for my experiment (E-94-016) which is staffed by personnel from Notre Dame, Indiana University, U. Connecticut, W+M, and U of Richmond. This makes it very difficult to get things done when we first show up to work. Everyone spends half a week getting training and paperwork straightened out before they can go to work. And it´s very hard to do any kind of remote work at all, because all procedures at (for example) the User Liason Office, or the Computer Center, are designed around the user showing up at the office. For example, to request new computer accounts, or to have something done on your computer account, requires that you physically show up at the computer center, but that´s impossible for researchers in, say Indiana University where I am. new management The increase of facility development time. Try to do fewer experiments better. Increase the computing staff to the point that software could be adequately tested and documented before being used to acquire data. increase the staff of the hall Maintain better and more up-to-date documentation of changes in running procedures and instrumentation More people, both JLab and university. Better overlap and communication between the group whose experiment just finished and the group whose experiment is about to begin. I would like to see lab management make a commitment to running good experiments, not just getting the experiments on and off the floor on some semi-arbitrary schedule. Getting a larger machine efficiency For Hall B, hire some software professionals who will not devote their time on physics experiments instead of what they are hired for. Get large inexpensive (Linux) IDE-RAID disks. Enough for each experiment to store ALL their data on the disks. This is a modern solution and quite feasible within a reasonable budget. Ian Birds comment of "These IDE RAID disks are so cheap, there must be a problem with them" sums up his attitude of expensive=good, which is one of the limiting problems with the computer center. See comment under 7. Improve the quality of life and education of graduate students located at Jefferson Lab and help in graduate student recruitment have a more coherent system for initiating new users. i have been at jlab twice, (once two years ago), and each time, it has taken me two full days of running around the lab to get my badge. this included going back to people once and then again to get signatures, and filling out the same personal information 3 times, even though i had already done it all two years ago. it would be nice to have a packet ready for new users with all the steps tied together. also, pre-registration on a web-page, with a direct link to the radiation training material to read in advance, would be very handy. I feel this is the wrong question to ask. The process of doing experiments is close enough to the right method so that only minor fine-tuning is needed. The ´single thing needed´ is to develop a vigorous and open science culture at the lab. Rather than have a focus on ´running experiments´ in a particular hall, what is needed is to cultivate a broad focus on scientific questions, where there is strong and consistent participation from both theorists and experimentalists, and both from the lab staff and the user community. This goes far beyond having the occasional seminar. I see little of this at the lab. A better balance between scheduling large experiments with special equipment and small experiments with standard hall equipment, especially in Hall C. Make a larger cryotarget group. This may be obvious but the more computing power the faster and more effiently analysis is. I sympathize with the idea that it is partly user repsonsiblity to provide computer time but I believe this can only be done at the later stages of analysis. Cooking for Hall-B can only be done optimally on site and this is the bulk of computing time needed. The PI´s should have more a say in the calculation of the actual beam time based on the approved proposal and the various renormalization factors. - Updating for all users (even retroactively) measurements and systematic errors in (for each Hall): * beam energy, * beam polarization, * beam and target position * spectrometers angle and momentum calibration. Only one "big" experiment scheduled per year in Hall C. The smaller experiments should be interspersed appropriately there. need to reduce cost of room at RESFAC Less competitivity, more communication between people For users to better understand how the accelerator works and what are reasonable requests of MCC. This can be accomplished by either taking the MCC OFT or by going to MCC and observing the day to day operation of the machine. It would help if all parties connected with conducting experiments would reflect the attitude that beam time is expensive. I think the lab should create "faculty" of physicists who have different duties then providing support to the users. There should be a group who job it is to maintain experimental equipment and to manage the setup and installation of the experiments - project management. Many of hall physicists have professional apsirations that are frustrated by the responsibility to hold the hands of the users. They do not like to do it (should they?) and are conflicted with their own intersts and those of the users. View Comments

maintained by webmaster@jlab.org