David Hunteris creating iOnTheSky - a revolutionary aurora forecasting system
Select a membership level
Get access to iOnTheSky Generation 1 footage, plus get access to exclusive news updates about iOnTheSky. You'll also get access to the exclusive iOnTheSky Development Facebook group, where you can discuss ideas with me (David) and provide feedback and have your say. I want you to be a part of the development. Your feedback and advice is valuable to me!
Along with the benefits of the above support tiers, you will be able to install the iOnTheSky app on your iPhone or Android phone prior to its official release.
Please note that the app development system allows for only a limited number of early-access users, so there is a limit to the number of people who can access this support tier.
You get the benefit of all other support tiers, and you also get to propose one possible Easter Egg for within the app or within the generation 2 camera housing.
In mid-2017, I traveled to Tasmania and began installation of the first generation of the iOnTheSky aurora camera network. This network of cameras has seen many auroras over the last year and a half; almost a million images have been taken by the cameras! This first generation camera network was incredibly cheap, but it's not ideal.
I am now working on a purpose-designed camera system (nicknamed "Gen 2" or "generation 2") to replace the earlier camera network. This new camera network will have a significantly more-sensitive camera, a Ricoh Theta V, which shoots in 4K and in 360-degrees/spherical. With these new cameras, we won't ever miss an aurora again.
But why do all this?
Aurora photographers like to look at remote cameras (such as iOnTheSky) to try and predict what will happen in the next few minutes. But I think I can go one better.
I am in the process of creating an aurora-specific convolutional neural network (CNN): software that will automatically detect various types of aurora in photographs.
Using this Aurora CNN, iOnTheSky will automatically create a database of dates/times when aurora was seen in each of the cameras in the iOnTheSky network.
Feeding this data, along with other spacecraft data (such as DSCOVR SWEPAM data) into a long short-term memory neural network (LSTM NN) should create a fully autonomous, highly-accurate aurora forecasting AI system.
I have teamed up with my friend, Brad Moore (who runs an app development business called Four Pi), to develop iOS and Android apps to present all this data (and aurora forecasts) to you, the end user. We are making good progress, and I am confident all of the above can theoretically be achieved. The main challenge, at this point, is funding it all.
Please consider becoming one of my patrons; your support will let me speed up development of this system (right now, I'm funding it as I am able to from my consulting work and work from part-time jobs, which are infrequent and unreliable). I need your help, please.