Supporting Hubble and JWST has spurred quite a lot of work in various computational fields. For instance, I’m using the AstroPy functions.
Another nifty example:
Although Artificial Intelligence and Discrete Optimization had known and reasoned about Constraint Satisfaction Problems for many years, it was not until the early 1990s that this process for solving large CSPs had been codified in algorithmic form. Early on, Mark Johnston of the Space Telescope Science Institute looked for a method to schedule astronomical observations on the Hubble Space Telescope. In collaboration with Hans-Martin Adorf of the Space Telescope European Coordinating Facility, he created a neural network capable of solving a toy n-queens problem (for 1024 queens).[3][4] Steven Minton and Andy Philips analyzed the neural network algorithm and separated it into two phases: (1) an initial assignment using a greedy algorithm and (2) a conflict minimization phases (later to be called “min-conflicts”). A paper was written and presented at AAAI-90; Philip Laird provided the mathematical analysis of the algorithm.
Subsequently, Mark Johnston and the STScI staff used min-conflicts to schedule astronomers’ observation time on the Hubble Space Telescope.
Animation of min-conflicts resolution of 8-queens. First stage assigns columns greedily minimizing conflicts, then solves
Min-Conflicts solves the N-Queens Problem by randomly selecting a column from the chess board for queen reassignment. The algorithm searches each potential move for the number of conflicts (number of attacking queens), shown in each square. The algorithm moves the queen to the square with the minimum number of conflicts, breaking ties randomly. Note that the number of conflicts is generated by each new direction that a queen can attack from. If two queens would attack from the same direction (row, or diagonal) then the conflict is only counted once. Also note that if a queen is in a position in which a move would put it in greater conflict than its current position, it does not make a move. It follows that if a queen is in a state of minimum conflict, it does not have to move.
This algorithm’s run time for solving N-Queens is independent of problem size. This algorithm will even solve the million-queens problem on average of 50 steps. This discovery and observations led to a great amount of research in 1990 and began research on local search problems and the distinctions between easy and hard problems. N-Queens is easy for local search because solutions are densely distributed throughout the state space. It is also effective for hard problems. For example, it has been used to schedule observations for the Hubble Space Telescope, reducing the time taken to schedule a week of observations from three weeks to around 10 minutes.[5]
Mollick collects some research about what’s known about people working in teams:
As team size grows with fixed task size, people individually put in less effort, but there’s wider collaboration, leading to better ideas
Teams that alternate between working together and time spent working individually come up with a broader set of possible solutions than either teams working entirely as groups or entirely solo
Collective intelligence of a team is limited by the emotional intelligence of the least sensitive team member
Teams that have good processes are more likely to have higher collective intelligence, because process let’s people asynchronously coordinate
Teleconference calls without video seem to promote more frequent syncing and have more even turn-taking than with video on.
Electron dynamics in complex time and complex space, thesis of Emilio Pisanty Alatorre
And, as a last call: I am grateful to the tram tracks near the Karl-Ziegler-Straße station,
for helping me find the time to finish the research presented here, albeit in slightly more
painful a manner than I might have liked.
The Maximum Principle of Pontryagin in control and in optimal control, Andrew D. Lewis
Like much of mathematics, one can come to some sort of understanding of the Maximum Principle by applying it enough times to enough interesting problems. However, the understanding one obtains in this way may be compromised by looking only at simple examples. As Bertrand Russell wrote, “The man who has fed the chicken every day throughout its life at last wrings its neck instead, showing that more refined views as to the uniformity of nature would have been useful to the chicken.
This is a stellar report - one of the best I’ve read in a long time. It has all the trappings of a good NTSB report. Something I really appreciate is how these usually span many layers; they discuss the local regulatory body drew a few
Write down every procedure, even if you don’t need them to be written down
Cross-check the outputs, not the inputs.
The verification QA checks on phantoms were performed .
There are a few different “types” of validation that you can write into a procedure.
Process step (Step 14: Close the pneumatic valve).
V.1 Did you complete the process step correctly? (was the right valve closed, or did the operator accidentally do the wrong one?) Check that valve #49 is in the “closed” position. (this may be redundant in many cases, but it also helps safeguard against interrupted work)
V.2 Did the action have the correct immediate effect? (Step 16: Confirm that the pressure in the pneumatic manifold has dropped) can also be used to . safeguards against failure modes like - the stop on the valve moved, the valve leaked, etc
V.3 Did the action have the correct overall behavior, and did that behavior have the desired outcome? ()
Even in my work, I tend to write checks in checklists ‘locally’. Let’s say I want to
Let me give you an example from my work.
Via a control panel, I set a certain meter from an auto-range mode to manual ranging. There was, however, a latent bug in the configuration panel.
I performed the check, but the check was in the wrong place. I was extremely lucky that the ammeter happened to be in the correct range.
The
The “Auto” light on the front panel of the ammeter, which was visible. I now perform this check every time I make this change
A behavioral test; inject a specified current and observe that the meter changes range correctly.
An all-up test;
In a sense, you want at least one part of your cross-check to “span the largest possible distance”; including all the intermediates.
Another example: some time ago, a procedure was followed. (MIP valve)
Save the data
While the inputs from the doctors were recorded, the outputs from the computer (the dose timings) were not recorded.
Figure 2 presents the structure of an MSA shutter. Each
consists of a door with an embedded electrode (wired to the ‘171-
side’ circuit) and etched magnetic strips, hinged at one side to
open into a crate-like housing which contains a second electrode
on the hinge-side wall (the separate ‘365-side’ circuit). Configur-
ing the MSA involves a coordinated electromagnetic procedure
which varies the electrode voltages while sweeping a moving
magnet arm back and forth across the shutter arrays. During the
outbound journey from the primary park position of the magnet
arm near the IFU aperture, in the dispersion direction (towards
the shutter hinges), the door and wall electrodes of every shut-
ter are set to opposite potentials, generating an electrostatic force
which latches the shutters in place once the magnet pushes them
open. After completing this first motion, the magnet arm reverses
direction and returns to primary park. Initially, all shutters retain
their latching potentials, but as the magnet arm passes over each
row in turn, the shutters in that row are individually addressed,
holding those that need to remain open (latched), while discharg-
ing the voltage for those commanded to close. Shutters with the
electrostatic charge removed are gently pulled closed by the pass-
ing magnet. In this manner it is possible to configure the MSA
into any combination of open and closed shutters.
The scientific case for a third-generation soft x-ray facility such as the ALS had always been fundamentally sound. However, getting the larger scientific community to believe it was an uphill battle
If somebody kept a very accurate record of a human being, going through the era from the Gay '90s, from a very different kind of world through the turn of the century—as far into the twentieth century as you might live. I decided to make myself a good case history of such a human being and it meant that I could not be judge of what was valid to put in or not. I must put everything in, so I started a very rigorous record.
— Buckminster Fuller, Oregon Lecture #9, p.324, 12 July 1962[3]
In public appearances, Fuller always wore dark-colored suits, appearing like “an alert little clergyman”.[80]: 18 Previously, he had experimented with unconventional clothing immediately after his 1927 epiphany, but found that breaking social fashion customs made others devalue or dismiss his ideas.[81]: 6:15 Fuller learned the importance of physical appearance as part of one’s credibility, and decided to become “the invisible man” by dressing in clothes that would not draw attention to himself.[81]: 6:15 With self-deprecating humor, Fuller described this black-suited appearance as resembling a “second-rate bank clerk”.[81]: 6:15
