Status Update #7
Photos for Getting Started guide, support for new water sensing use cases, and real-time alerts
It has (again) been a little while since the last update.
The plan had been to follow up fairly quickly after opening Early Access, but the past few weeks have been split across a number of different priorities. Some of that has been progress on the project itself. Some of it has simply been life catching up.
This last month has brought an onslaught of distractions that have been stealing more time away from this project than I would like. Our 25-year old water heater decided it was time to give up the ghost. Spring chores on the farm are picking up, and Uncle Sam owed me some money this year, so I’ve been prioritizing getting that wrapped up early. Somewhere in the middle of all that, I also “accidentally” picked up a small bandsaw mill from the local auction, which was not exactly on the todo list but has been on the bucket list for quite some time.
Early Access Update
Even with the limited time on HTH, progress has continued in a few parallel areas.
On the Starter LoRa side, I have been working through updates to the Getting Started Guide that is currently out for early access. The main focus here has been completing the photos to match the walkthrough steps and give a bit more visual structure to the guide. The overall flow and sequence are already solid, but the guide was really heavy on text. The goal is to make it easier to follow without needing to read quite as carefully on the first pass.
Expanding Simple LoRa
Most of the recent engineering time has been spent on the Simple LoRa tier.
The biggest improvement here has been around downlink reliability. Earlier versions of the firmware would occasionally miss packets, especially when multiple nodes were active. After some investigation and adjustments to the radio handling, that behavior is now much more reliable. Downlink commands are consistently reaching nodes, even with several devices transmitting.
I have also started assembling the first pieces of what will become a water monitoring and irrigation-focused sensor set.
The initial plan is to cover several different sources of input for water
A pulse-based tipping bucket rain gauge and/or pulse-based water meter measures rainfall and applied irrigation
A Vegetronix VH400 soil moisture sensor measures water infiltration and retention in the soil
Initial work toward integrating higher-end soil-sensors including the TEROS-10 and TEROS-11/12
Pulse-based counter support is implemented and I have been successfully logging data through the spring rains with the tipping bucket rain gauge. I still need to test this with the pulse-based water meter, but I expect it to behave similarly with some slight adjustments to account for calibration differences.
Both the TEROS-10 and the VH400 are analog moisture sensors that require calibration to produce accurate measurements for volumetric water content. I have implemented initial support for reading the raw voltage output of the sensors but still need to sort out what it looks like to convert that to a VWC percentage.
I have a TEROS-11 unit on order with the intent of adding SDI-12 support for the digital style sensors that come pre-calibrated from the factory. That will open the door to a different class of sensors that are more commonly used in larger data-logging applications. TEROS sensors are quite a bit more expensive than the hobby-grade moisture sensors on Amazon (like 50+ times…) but it seems they are far more accurate and reliable when exposed to the elements. The VH400 stands to be a good compromise on price vs. reliability but more testing is needed to confirm this. I’ll be deploying a couple of these here soon and will report my findings here.
At this point, most of the individual components are either working or very close to it. The focus now is on getting them to behave consistently and over time.
Toolkit Updates and Alerting
Alongside the firmware and sensor updates, I have been working on the first pass of a rudimentary alerting system.
The idea is that you can define simple conditions based on sensor data or device status, and the Toolkit will push a notification to your phone when those conditions are met. This includes things like a sensor going offline or a measurement crossing a threshold.
This feature does require internet access to deliver the notification, since the message is passed through a lightweight 3rd-party notification service. The rest of the logic still runs locally in the Toolkit. Only data that you include in the alert payload is sent to the external service. The initial alerting system will be built on ntfy.sh, which is an open-source and auditable project and can be hosted on a local server if desired. That is out of scope for this initial push, but will likely be a topic for a future post.
For those that prefer to keep everything entirely local, alerts are completely optional and opt-in. Enabling/disabling does not affect the core system functions if you choose not to use it.
What’s Next
The immediate priorities are fairly straightforward and should all be doable with some focused effort over the coming weeks.
Finish the photo updates for the Starter LoRa guide
Get the water monitoring components working together as a cohesive setup
Add support for the remaining sensors, including SDI-11/12
Continue hunting bugs throughout the system under real use
If I can get all of these things to come together as I hope, the next step will be packaging this into something that can be shared more broadly.
Overall, progress this last month has been slower than I would have liked, but it is still moving in the right direction.
If you want to follow along as things take shape, consider subscribing.