Here’s a project idea that comes up occasionally and would be useful. I don’t have time to do it, but would help and support someone who wants to take it on!
Background
Existing Laser Telemetry
We have a laser telemetry sensor box that’s been running for about a year. It can track laser power and usage, as well as chiller temperature and filter pressure differential. It stores this in Grafana where it can be graphed and stuff. This is how @JDMc makes those data-sciency graphs in the Laser Cutter Committee Reports
Laser access / lockout
There is a raspberry pi beside the laser cutter. Users have to log in with their account and then flip a switch to turn on the laser. Everybody who has done laser training knows what I’m talking about. The Pi runs this nodejs application which has not been touched in 4 years (but works perfectly).
The idea
The laser access software knows which member turns on the laser, and it should be possible to write that to the same InfluxDB as the other laser telemetry. Perhaps for privacy reasons, it makes sense to write a pseudonymous identifier instead.
Then it would be possible to make a report to show this distribution of laser users and how much they each use it. To generate insights like:
Last month XX different members used the laser
Last month, the top 5 laser users accounted for XX% of its overall usage
So,
If this is an interesting project to you, I’d be happy to help get you going. If you get comfortable with the laser access software, you could also clone/extend it to support other tools at VHS.
There’s already a ./last-laser-users.sh in the home dir of the pi user on the laser-pi.
Currently that information is logged both on the Pi and Nomos logs it in its access logs.
AFAIK, LCC has access to the laser-pi, so, it might be fairly easy to take the last-laser-users.sh output and correlate that to actual use.
That said, if anyone wants to contribute code to the ATOMS project for possible user-usage telemetry, the help is more than welcome. ATOMS is meant to replace the current laser-access tool and provide infrastructure for all tool-lock systems at VHS.
@LCC: if anyone of you doesn’t have access to the laser-pi or if the password has collectively been lost, let me know, or else if anyone on the LCC wants output of the last-laser-users.sh, I’ll be happy to provide this.
Thanks for the responses Ty! We already have access to the raw data, or at least the log files on the pi. I didn’t know about last-laser-users.sh (I normally just tail the log file) but that sounds like it could be useful. It makes sense that Nomos knows when the auth happens, but I assume it doesn’t know when the user logs out of the laser pi. So the laser Pi’s data is probably more useful here.
If you’re offering to turn that data into insights like the ones in the OP, then that would be awesome! Ideally it should be done in a way that can be run monthly, and I think writing the data to Influx (in addition to the log file) might make that easier, but even a one-time analysis is better than nothing.
But also are you sure you want to take this on? You already do so much IT work around VHS! I was kind of fishing for a newish member to step up and use this as a tidy little project to start getting involved. That’s the ulterior motive of a “someone should” post.
I wanted to see this through, so I gathered the logs off the laser pi and wrote a script. The following is a list of users ordered by the cumulative usage since May 16, 2018 (except Sep 1 — Oct 14 for which logs were missing). So this is a sampling of nearly 6 months.
Logged in
Laser energy
Sessions
User ID
5 days, 3:29:06
1 day, 19:12:07
49
323
2 days, 13:22:50
20:57:58
30
393
2 days, 6:21:29
7:12:31
43
398
2 days, 5:57:21
9:03:03
28
350
2 days, 3:32:23
6:06:21
27
365
1 day, 0:40:50
3:50:11
15
409
23:54:24
2:35:29
20
257
14:37:11
3:34:36
6
21
12:03:39
2:52:20
7
480
11:05:30
0:50:13
12
500
9:29:47
0:49:19
7
404
9:01:58
1:21:29
8
427
6:55:57
0:05:08
7
453
6:32:46
1:15:30
4
56
6:09:28
0:57:59
4
540
5:53:43
0:07:06
2
505
5:42:47
1:07:14
27
375
5:32:22
0:19:03
2
264
4:50:55
0:31:05
8
495
4:47:53
1:12:25
3
400
4:13:29
0:30:55
8
504
3:50:45
0:41:10
2
212
3:16:36
0:12:48
4
133
3:11:11
0:13:04
1
75
3:08:58
0:32:54
3
515
3:04:20
1:27:27
2
502
2:50:43
0:50:45
5
202
2:47:49
0:06:46
1
346
2:36:22
0:21:30
9
164
2:11:09
0:01:16
7
292
2:07:05
0:36:16
3
471
2:02:01
0:06:39
5
116
1:46:32
0:03:55
3
429
1:42:54
0:00:47
2
128
1:29:32
0:14:33
1
498
1:12:35
0:06:15
1
312
1:08:11
0:01:06
1
89
1:08:05
0:04:30
1
550
1:07:59
0:10:42
2
101
1:01:09
0:12:12
4
247
0:59:47
0:07:37
1
510
0:55:40
0:15:56
1
491
0:52:22
0:00:53
2
332
0:32:00
0:00:26
1
543
0:31:10
0:08:14
1
120
0:15:46
0:02:10
1
516
0:14:05
0:02:41
1
170
0:09:28
0:01:53
1
501
0:08:48
0:01:16
1
507
0:03:02
0:00:00
1
436
What the columns mean
Logged in is the time that the laser system was logged in by each member, whether the laser was energized or not. This gives a sense of how long the laser was tied up by each member.
Laser energy is the cumulative time that the laser fires at its full power, which on our laser is “60%” in LaserCAD. This is sampled at 1kHz. If you cut something at only 30%, then this number accumulates half as quickly, and so on for other power levels. This gives a sense of how intense the laser usage was for each member.
Sessions is the total number of unique sessions. This shows whether a member had a few long sessions, or lots of short sessions.
User ID is your Nomos user id. To find your own userid, log into Nomos, and then visit this link.
A reminder that our current policy allows unlimited use of the laser, so my aim is not to find any fault in these usage patterns. In fact I’m pleased that some users are getting such good value from the laser! My goal in collecting this data is so we can make more informed policy decisions going forward.
Thanks for doing this, especially without adding stigma to it! I love that the laser is so useful to so many people, my own desire is that this data helps serve as direction for the future, with no expectation of what measures it will influence. Thanks!
