I wanted to start a topic specifically for those of us who either have an observatory or are thinking of building one be that remote robotic or not. Whether you're in the planning stages or already running a fully automated system, I believe we can all benefit from sharing some of the lessons we’ve learned along the way.
I operate a fully remote robotic RoR observatory, and while it's currently stable and productive, it certainly wasn't without its learning curve (a year of testing everything). I’ve been through several iterations and improvements both in hardware, software and workflow — and I’m still optimizing processes, automation scripts, and organizational systems. Especially optimising the building managment system via Zabbix and Grafana. (I will eventually make a seperate post specific for my obs but in the meantime this is much more educational for everyone reading this)
The goal of this post is simple:
Let’s share the things we wish we knew at the start, the mistakes we corrected, and the “a-ha” improvements that made a difference. Big or small, your tip might save someone else a major headache or help make their setup more robust.
A few lessons from my own experience to kick things off:
1. Redundancy and remote control matter more than you think.
Having a smart PSU/UPS and IP-controlled power for everything (including your network gear, USB hubs, and even the roof controller) means I can recover from 90% of issues without having to physically visit the site. Redundancy is key especially remote, as of writing this I run N+1 (one on-site spare) of all non-critical items (every cable, connector, controller, rpi, arduino, sensor..) with ideally for almost everything. If it can fail remotely, plan for how you’ll fix it remotely too.
2. Weather monitoring needs layers, not just a single sensor.
Multiple sensors from multiple manufacturers. You don't want to rely on 1 sensor for everything. This helped me avoid both false positives and dangerous situations.
Use of smart home sensors was a really cheap solution from some of my implementations with already existing reliable software and a device with an affordable price.
3. Automate your logs and backups early.
From FITS files to mount logs and system health data — automate all of it. Having logs that sync offsite helped me catch a mount ASCOM issue before it became a catastrophic failure. (Yes my mount did try to kamikaze itself on some weird instances, and you should always be aware that bugs happen even if you think its perfectly set-up)
4. Label everything. Document everything.
Future-you (or someone else) will thank you when troubleshooting six months later. I now keep a Google Doc with every connection, IP address, software version, and config change. Every change gets timestamped. Everything gets logged on a cloud (google or whatever is your pick), and a physical copy in form of instructions and labels on A4 paper on-site.
5. Everything needs to be idiot proof.
This follows my 4th lesson... Yes, most of the people running their observatories or plan on running them are usually well versed in the software and hardware side of things. However, there will be a moment where something needs to be unplugged or clicked or changed from an external user (be that a friend, family member etc.), everything you do write in a "manual" should be idiot proof. At least have 1 copy for the "public".
6. Look what others do and IMPROVE
Basically what this thread is about, before I built and automated everything I spent few months doom scrolling every forum post, cloudy nights post etc. looked at 100s of images, plans, sheets, documents. Saw what was problematic and what should just be copied and pasted into my version. It helps a lot and you should not start from 0, there were people already that struggled and paved the way so me and all of you can follow what is great out there.
(Look at professionals also, yes their solutions are probably more expensive but some can be very well implemented, it did help me that im working at such place but lots of data - such as images are public. And dont forget to reach out, everyone is on the internet nowadays, I can find a system administrator of multiple observatories within an hour, email and ask questions and their solutions...)
7. Have fun whilst you are doing it
Also, there are some excellent external resources worth checking out from great people on AB:
- John Hayes’ presentations on remote observatory automation — great insights for both beginners and advanced users.
https://www.astroworldcreations.com/blog — fantastic blog with observatory build, automation ideas, and tools.
[*]There are plenty other more, but I shared what I personally liked the most.
Please feel free to add your own lessons or thoughts — whether it’s something simple or a major breakthrough. Let’s make this topic a useful reference for anyone heading down the observatory path remote or not.
Clear skies,
Luka
