Vera poses a fun question to think about over the holidays! It
seems to me two areas are most likely to effect NMR over the next
decade or two, but I'm likely only smart enough to see that they
will change things but not to predicting very accurately how.
The Internet is an obvious point of focus, but here is where I am
most uncertain as to the impact. Will future researchers FedEx
samples to big centers and run experiments remotely (or have the
instruments run everything for them), or do people just have to
have the physical connection to the equipment and experiments?
I think the former is likely, at least for the biggest equipment.
It seems fairly clear that the Internet is becoming a focal point
for training, instrument guides and manuals, and user interactions
similar to this group. And although I recently recommended against
remote installs from vendors, remote servicing works quite nicely.
I do not see the Internet conferences replacing traditional
conferences, as I and I think the younger generations will require
the direct interactions of traditional conferences.
In my mind, the biggest changes will come about from experiment
automation. I definitely believe the instrument will become smart
enough to perform all experiments in an automated fashion; we are
already close to achieving this. It seems more of a stretch that
the instrument will decide which experiment(s) to run, although
perhaps for certain classes of compounds. But it seems clear to
me that separation between users and facility staff/spectroscopists
will increase to a point where the normal user (e.g., synthetic
organic chemist) is completely dependent on the staff to keep the
spectrometer at a usable state (mostly from a software perspective,
but including pulse sequences and pulse width calibrations, etc.).
The idea of large automated centers, even if they are sufficiently
distributed such that users can physically always be present,
has the potential of having weird parallels to large mainframe
computer centers, where a bunch of NMR geeks (us?? :) that cannot
seem to talk in a language users can understand (acronyms/spin-
physics already doing this) run around keeping the hardware and
software going, and performing seemingly esoteric experiments
that have little audience beyond our own peer group. The
spectrometers will have highly automated software diagnostics and
run regular spectrometer tests that automatically check with
vendor databases to insure all is running correctly. Users come
in and request the experiments they want (or perhaps just a protocol--
ie. need conformational check of the following structure) and the
spectrometer takes it from there. Results are posted directly
to the users electronic in-box. How much NMR, especially
about the hardware and experimental setup, the user needs to know
is in my opinion very much in doubt, although that very likely
will depend on the type of research being performed.
Seems to me the biggest difference between academia and industry
is that industry will go much stronger toward total automation,
whereas academia will have "ivory towers" of spectroscopists, and
people and equipment to train and cater to this group. Here at
UW, I already am considering a true separation of the facility
directed at supporting (experiment) automation for synthetic
chemists, and another portion of the facility providing training
and tools for our next generation of spectroscopists. I really
do not clearly understand the implications of doing this, but
see the diverging interests of my user base: one that needs
to obtain complex data with minimal involvement, the other
needing to learn the fundamentals of experimental NMR. This is
an issue that I am hoping various of you are already thinking
about (i.e., how much experimental automation to use); I will
be lead a discussion section on this at the AMMRL section of
the ENC in Orlando.
Looking over what I've written so far, it seems apparent to me
that I am not looking ahead far enough. We are already on the
edge of doing the things I suggest above. Database and
theoretical developments could provide automated interpretation,
and perhaps smart/automated experiment selection. Will software
be smart enough to do a full unknown analysis without help? Will
users in 20 year not know what a ppm is, or what a J-coupling
is, as the computer is doing all the interpretation for them?
Is a Star Trek-like system of ask the question and the computer
spits out the answer going to evolve? I don't know... much of
this is just a bit too far a stretch for me to envision. But I
can see the day coming when the student coming in not knowing
anything about the instrument or a pulse width or even what a DEPT
or COSY experiment is might be the norm, rather than just the
troubling times--seemingly way too often--when that happens now.
Best wishes,
Charlie Fry
At 09:21 AM 12/14/98 -0600, you wrote:
>Hello to all!
>
>As we settle into finals week here at the UIUC, I thought I would post a
>question that seems appropriate as we approach the new year and new
>millenium. Actually, it was brought to me by a colleague, and as I don't
>often think reflectively about where NMR as a field has been and where it
>is going, it seemed appropriate to post here. Specifically, where is the
>field going in the next 10-20 years? What new data will we be able to
>acquire? What hardware and software will be necessary to get there?
>
>When I was asked this question it was posed along with the question of
>advances in artificial intelligence and how that would impact data
>collection and analysis. I answered that I believed that artificial
>intelligence systems could and would have a significant impact in the next
>10-20 years, but that in my realm, that of training the next generation of
>users of nmr instrumentation, I would probably never take full advantage of
>such software. My reason was that having worked in industry, and also in
>academia, I believe that academia should train in the basic skills, e.g.,
>the importance of parameter choice in taking data or apodization for a nD
>data set. It is possible that in 10-20 years an nmr user will truly not
>need to understand anything, but let the instrument make all the choices.
>I believe that is a dangerous precedent in initial exposure to an instrument.
>
>So, that's my $0.02's worth (and I didn't even realize until now that the
>"cent" sign no longer exists on my keyboard - obviously I never use it!).
>I will post a summary, as usual.
>
>Vera
>Vera V. Mainz
>School of Chemical Sciences email: v-mainz@uiuc.edu
>142B RAL, Box 34-1 phone: 217-244-0564
>600 S. Mathews Ave. fax: 217-244-8068
>Urbana, IL 61801
>
>
>
----------------------------------------------------------
Charlie Fry Tel: (608)262-3182
Director, MR Facility Fax: (608)262-0381
Chem. Dept., Univ. Wisconsin
Madison, WI 53706 USA email: fry@chem.wisc.edu
----------------------------------------------------------