In a previous life, back in the pre-CCD days I worked at an observatory searching for new asteroids. Used to be the way it was done you would point a telescope

In a previous life, back in the pre-CCD days I worked at an observatory searching for new asteroids. Used to be the way it was done you would point a telescope at the part of the sky directly opposite the sun (min r^4 so best chance of detection), expose a photographic plate, and repeat a few hours later. Develop the plates, mount them in a blink comparator [1] and scan manually, flipping back and forth between the plates for ~3 hours to look for anything that moved. Exclude false positives like plate blemishes (aka Eastman-Kodak Objects) or dust grains. If you see something, and it is smeared indicating motion during the exposure, mark it as a possible candidate. Revisit the same patch of sky the next night, search again. Use the datapoints to extrapolate forward, check again a couple of weeks later. If you haven't lost it (which frequently happened if it was at the limit of visibility), extrapolate back to the previous opposition (~4 years for a main belt asteroid) and search the archives for a match. If you get one, then extrapolate further back and check all previous oppositions. If you get a match on three separate oppositions, congratulations, you have a discovery. Vastly different now - a single observation can net large numbers of asteroids and sensitivity is so much better. Still the search for this potential new planet is definitely going to be daunting given the distance. [1] https://en.wikipedia.org/wiki/Blink_comparator https://en.wikipedia.org/wiki/Blink_comparator