
15,000 light years they say
Those double glazing adverts are relentless.
Astronomers believe they may have discovered a new type of stellar object after spotting something that has been beaming radio pulses every 22 minutes for more than three decades. The oddity, code named GPM J1839−10, lies about 15,000 light-years away, and was detected by an international team of researchers. They were …
Another "hmm, that's strange" moment in scientific history that will likely produce a major discovery and, possibly, new mechanics in astrophysics.
I'm rather excited about this, can't wait to hear about it from scientific sources.
That said, the data has been stored for the past three decades, and they only find it now because nobody was looking for it before ?
Somebody get an AI on that data, pronto. For once we have something a statistical analysis machine can actually be good for, it would be shame not to.
Einstein@Home is a similar project, trawling radio telescope data for undiscovered signals, although they're more interested in binary pulsars to find candidates for gravitational waves.
Considering all of the astronomical data that has been collected over the decades and not processed at the time, because of cost or a lack of something or other (computing speed, software, etc ?), it is good to see that the recorded petabytes/exabytes/or-larger (quettabytes?) of data are being re-examined and yielding new information and understanding...
As long as the input data is not garbage (unlikely?) there will not be garbage out...
Using a computer-learning model (I hate the use of "AI") to analyse ALL of this old data will probably lead to even greater understanding of the Universe we live in and the Planet we live on
Please don't cry: much of the old data ROTted.
The data was on magnetic tapes. They were not all cared for properly. IT started noticing a phenomenon called ROT. You had a Read Once Tape. The tape is made of mylar with ferro magnetic metal glued on. When one of these tapes is read, the bending of the mylar around the read heads of the tape drive is enough to cause the deteriorated glue to release the metal oxide. Once the tape has been through the drive once, much of the data has literally fallen on the floor. So they developed a procedure for thoroughly cleaning the tape drive, reading the tape through and capturing what they could, and trying to rebuild lost data they could infer from checksums. It was a tedious process, and for data that was deemed worth it they could get quite a bit off the tapes. Not all the data was recovered, and not all the data was deemed important enough to spend the budget on trying to recover it.