Althea Development Update #44: Picking up steam

Althea now has 4 full time users with more in the immediate pipeline and longer term plans shaping up.

Two are using CPE’s as pictured below for long distance links, of a few yards to a mile. The other two are connected using the 5ghz wifi in the routers themselves, which is performing flawlessly over short distances with only 3 peers.

These are representative of our two major use cases, high density housing where short distance ‘pure mesh’ is possible and long distance communication using specialized hardware. In an apartment complex for example a home router would be able to resell to adjacent units above and below. Then specialized hardware would be used to reach beyond that.

The key leverage of mesh technology in Althea is not to create a network without infrastructure, but to create one out of any infrastructure available automatically.

Keeping up with the influx of new interest is really cutting into development time for me. It’s fortunate we already have one employee up to speed, but as the resident cross compiling and tooling expert our on router tooling and specifically process improvements for new deployments are a high priority for me.

The faster I can start posting ready-to-use firmware files on our website the quicker users can move to flashing their own routers without my intervention. Which is currently required not because there isn’t automation, but because that automation is too environmentally delicate for a non-expert to setup.

The router software now has auto setup features to fix this. So it’s a matter of getting it properly integrated into the firmware and our production tooling. We also have the first iteration of our billing software code complete, so it’s undergoing integration testing with the auto setup software.

One of the big challenges is updating core network software on remote devices in such a way that they will come back online and remain compatible with the rest of the mesh. Since a split-version mesh without interoperability will isolate devices that need the mesh to download the update.

These catch-22’s require manual human intervention, so we’re going to take even this first update cycle pretty carefully.

It feels good to be making steady progress. Especially since we keep seeing our engineering decisions pay off.