# ISRO's PSLV-C37 with 104 satellites - after 56 hours only six TLEs, which factor is greatest challenge?

It's been 24 hours 56 hours since ISRO's PSLV-C37 launched from Satish Dhawan Space Centre (SRI) at 09:28 local time (03:58 UTC).

I see satcat numbers and object names for 105 total objects; 41948 2017-008A through 42052 2017-008DJ. It looks like only six deployed satellites have TLEs so far A-F with DJ appearing to be the fourth stage of the PSLV rocket still in orbit.

24 hours 56 hours is longer than I expected for first TLEs to be released.

I'm pretty sure this is quite a challenge, and I can think up a number of them here:

• n > 100
• low optical cross-section
• no telemetry or on-board GPS fixes reported from satellite owners

Do any of these (or other) factors stand out as a primary reason why this would be so challenging? Are any of them not really a reason?

Videos below: P- and P+ views appear to be nadir and zenith cameras.

These are pretty darn cool videos - but check that your volume is not at maximum before playing. Vids similar but not the same, waiting for a single definitive video to post instead.

Also available as a Planet Labs Tweet!

• It's nice to see such a deployment not coming from the belly of a B-52 for a change, there's hope for us yet. – uhoh Feb 16 '17 at 5:13
• Those sizes are right at the edge of reliable trackability using unclassified assets. Getting accurate TLEs for a 10 cm object 500 km away is going to be tricky at best. They're barely visible. – Tristan Feb 16 '17 at 18:14
• @Tristan I'm pretty sure 88 of them will look like this soon after deployment i.stack.imgur.com/01vtm.jpg While Planet labs keep careful track of of their Doves' GPS telemetry (fourth item on my list) doves do also get TLEs. These are 3U's with unfolding panels, even 1U's get TLEs don't they? OK let's find out! – uhoh Feb 16 '17 at 23:35
• @Tristan I've just asked Are 1U cubesats sufficiently detectable to get at least minimally usefully predictive TLEs?. – uhoh Feb 16 '17 at 23:51
• If you're interested, Planet publishes their own TLEs, state vectors, and JSpOC "associations". – Chris Feb 17 '17 at 15:17

From experience, I assume that Space Track is in fact tracking them all. The difficulty lies in making positive identifications of the names of all of them, which will happen soon. As the separate further, this identification will be made easier, but I'm confident that in the end, all such objects will be listed, and all of them are being tracked.

FYI, this is standard procedure. Falcon 9 launches with less than a dozen satellites typically take days for their TLEs to be released to the public. This is nothing unusual. I don't understand the process completely, but I believe that senior personnel at JSpOC have to review the identification before it is released to the public.

• I'm sure your right, and when things settle down and any remaining potential ambiguities are resolved, the public databases will show TLEs. But assigning names does not seem to be necessary. Objects 41948 through 41953 (2017-008A through F) have had TLEs for over a day now in the public database, though only one of them has been given a name so far (2017-008E = FLOCK 3P-30). It's a fascinating system - thanks for your help! – uhoh Feb 17 '17 at 15:02
• Trust me, if the owner of one of the satellites calls JSpOC and asks for information about the TLEs, they will get more than they are currently listing. It just takes some time to process the identification of everything. They don't even want to assign a letter until they are fairly sure of what it is, for instance. – PearsonArtPhoto Feb 17 '17 at 15:07