When I worked at NASA Glenn Research Center in Cleveland 5 or so years ago I was also a researcher on this project - the previous Pettegrew/Struk paper referenced in the article was my team - I replaced R.S. Downs.
Interesting stuff really - our work mostly focuses on the fact that in space, buoyant forces are strongly squelched due to the lack of gravity, and so air voids in the solder never rise/convect out. They stay put and affect the strength of the joint itself. My role was to write some mathematical algorithms that could take the 3D images of CT scanned solder joints created in microgravity as inputs and return analysis of the percent voids, spacial distribution of voids, etc. They were all broken down into hundreds of 2D image slices, and automated algorithm I created attempted to discern voids from other things, etc. Good times, brings back memories!
Originally posted by !~TONY~!Great work for sure! So what caused you to leave NASA? I got belted out, so to speak, from Goddard when I was on the Apollo program, Apollo Tracking and timing. Heady times! It turned out the guy next door where I lived in Alexandria was also on Apollo, a geological tech, he cut the moon rock samples so the scientists can analyze them. He invited me to his lab and showed me the fort knox looking bank vault that held the moon rock samples and he picked one up and handed it to me. It was like an epiphany for me! I was awestruck to actually hold a moonrock in my hands! My first question to him was, 'why are you letting me hold this precious rock in my hand, I am contaminating it!'
When I worked at NASA Glenn Research Center in Cleveland 5 or so years ago I was also a researcher on this project - the previous Pettegrew/Struk paper referenced in the article was my team - I replaced R.S. Downs.
Interesting stuff really - our work mostly focuses on the fact that in space, buoyant forces are strongly squelched due to the lack of gravi ...[text shortened]... m I created attempted to discern voids from other things, etc. Good times, brings back memories!
He said, just taking them out of the vacuum pac box contaminates them exposing them to our atmosphere, so we don't really worry about the surface that much, the real story comes out when we slice the rocks into small slices that we worry about contamination. That is a different story.
That rock certainly looked alien! It was like no other rock I ever held in my hand for sure!
I wanted to continue on with my PhD (I got my MS there) but my adviser couldn't guarantee full funding/stipend, etc. In addition, the pace of the work and the long commute eventually wore me down, and I was ready for something new.
Very cool story about your neighbor - that's a once in a lifetime experience!
Originally posted by !~TONY~!What's your Ph.D.?
I wanted to continue on with my PhD (I got my MS there) but my adviser couldn't guarantee full funding/stipend, etc. In addition, the pace of the work and the long commute eventually wore me down, and I was ready for something new.
Very cool story about your neighbor - that's a once in a lifetime experience!
Originally posted by !~TONY~!One time during a training exercise on the use of the radio telescope/deep space network dishes, we used a small dish, I think about 12 feet diameter on the training building at Goddard. So my assignment was to find Mars, and as it turned out, there was an orbiter around Mars att, so my exercise was to lock on to the signal. Just lock on was enough. So I had to figure out where Mars was in the correct co-ordinates and aim the dish on the roof to that exact co-ordinate and then lock on to the data stream coming from that probe. I did it exactly right and got the 'locked' on the phase lock loop light! I am a ham and was thinking about that. I used a 12 foot dish to lock on to a Mars probe about 100 million miles away att and the probe RF power was 5 watts. Of course it had a parabolic dish directing energy to Earth but still, I'm thinking, I just communicated with Mars using a transmitter the power of a CB radio! A bit higher frequency of course🙂 But it was mind boggling nonetheless!
I wanted to continue on with my PhD (I got my MS there) but my adviser couldn't guarantee full funding/stipend, etc. In addition, the pace of the work and the long commute eventually wore me down, and I was ready for something new.
Very cool story about your neighbor - that's a once in a lifetime experience!
Originally posted by sonhouseThat is pretty extraordinary stuff! At my current job I have started doing some elementary RF modeling of antennae, and I find it fascinating. I have a very, very rudimentary understanding of it, but the physics of it really blows my mind.
One time during a training exercise on the use of the radio telescope/deep space network dishes, we used a small dish, I think about 12 feet diameter on the training building at Goddard. So my assignment was to find Mars, and as it turned out, there was an orbiter around Mars att, so my exercise was to lock on to the signal. Just lock on was enough. So I ha ...[text shortened]... he power of a CB radio! A bit higher frequency of course🙂 But it was mind boggling nonetheless!
Originally posted by !~TONY~!PhD's do tend to be an exercise in endurance. I got mine in the end but it took 5 years which ruled out the academic job I'd hoped for at the start. I reread your post and get it now. Was it blob detection? I implemented a connected components algorithm for a company I worked for using the two pass algorithm, one of the bits of code I'm more proud of.
I never got it because of the aforementioned problems - I have a B.S. in Mechanical Engineering and an M.S. in the Fluid & Thermal Sciences. 😀
The actual detection was something more simple in that we were overwhelmingly concerned with the aggregate void fraction of the entire joint. It was easy enough to threshold/segment the image and simply count 1's and 0's, then move on to the next slice and do it again. The tricky part was that the actual through-hole electrical component wires were also in there, so I needed to find a way to get rid of those in the analysis as well.
I eventually messed around with more sophisticated algorithms that tracked blobs through the void to try and track them & deduce things like overall shape using some fancy background subtraction algorithms, but that was mostly just for fun - I never used them for anything.
At my current job I've used two pass connected components analysis as well, and also ended up using some optical flow algorithms for object tracking, which was interesting. It's funny - I have no real technical training in Image Analysis/Processing, so I could be doing things really, really poorly, I dunno!
What do you do, if you don't mind me asking?
Originally posted by !~TONY~!Sorry for butting in, but do you know about these Comsol modeling programs?
The actual detection was something more simple in that we were overwhelmingly concerned with the aggregate void fraction of the entire joint. It was easy enough to threshold/segment the image and simply count 1's and 0's, then move on to the next slice and do it again. The tricky part was that the actual through-hole electrical component wires were also in ...[text shortened]... be doing things really, really poorly, I dunno!
What do you do, if you don't mind me asking?
http://www.comsol.com/models/rf-module
Originally posted by !~TONY~!I don't have any technical training in either programming or image processing either, the point of education is to be able to work this stuff out on the fly. The important thing is to make sure there's a good test program to make sure it's right. Then pass the code over to your infinite team of monkeys bashing a keyboard 😉
The actual detection was something more simple in that we were overwhelmingly concerned with the aggregate void fraction of the entire joint. It was easy enough to threshold/segment the image and simply count 1's and 0's, then move on to the next slice and do it again. The tricky part was that the actual through-hole electrical component wires were also in ...[text shortened]... be doing things really, really poorly, I dunno!
What do you do, if you don't mind me asking?
My thesis title was "Polymerisation of 2D Quantum Gravity Models", but these days I'm a computer programmer. The company I was working for at the time was a software company, there wasn't a specific project the program was meant for, we were implementing a standard set of algorithms called vsipl - vector signal and image processing library [1] - for some reason although the vector and signal processing part of vsipl is well defined and solid, the part of the standard defining the image processing functions is only at the first draft stage and full of mistakes so I had to unscramble their off-by-one errors. I'm freelance at the moment, or self-unemployed as a friend calls it - the project I'm working on at the moment is an epidemiology model of cycling patterns (health and transport) although this one is more about transport than health. Luckily for me, although they can produce a perfectly good program, their critical code used an N² algorithm where an order N algorithm was possible, so I managed a speed up of a factor of 7 (more for larger populations) which looked really good without being that hard to do, I also found a bug that was sometimes causing it to go into an infinite loop. So it's quite a nice project.
[1] www.vsipl.org