The first probe
Our first bench probe reads soil at two depths so we can see what watering reaches.
We put our first probe into a bucket of soil this week. The electronics sit beside the bucket, connected by wires that catch on everything. There is no sealed body, solar cap or phone app yet. We have a board, two moisture pads and a temperature sensor, and we can record a set of readings every 10 minutes.
We started because we kept making the same watering mistake in our vegetable beds in Halden. A dry surface sent us for the watering can even when the soil below was still damp. After a short shower, the opposite happened: the top looked wet, but digging down showed how little water had reached the roots. We wanted a way to watch both layers over time.
Why two depths
Our pads sit at 5 cm and 15 cm below the soil surface. The upper pad should show a small watering or a brief shower first. The lower one should help us see whether that water has moved farther into the bed. Neither depth describes every root system, and one probe cannot represent an entire garden. We chose these positions as a useful starting point for vegetable beds.
On the bench, we add water slowly and compare the readings with soil we can touch and weigh. We also leave the bucket alone between sessions. We need to understand the changes during drying before we can decide what the app should show. A number that rises when we pour water is only the first check.
The temperature sensor gives us another piece of context. Cold soil and warm soil behave differently, and the electronics can drift with temperature too. We are keeping the raw records rather than turning an early measurement into a watering instruction.
What needs to leave the bench
Our intended Stake body is 28 cm tall. That gives us room for the two sensing depths, electronics above the soil and a cap that we can reach without digging. We are drawing a glass-filled nylon enclosure, but the bucket version tells us very little about sealing that enclosure.
We still have to solve power, moisture getting inside and a radio connection that works outdoors. The screw-on solar cap is a sketch at this stage. We want a free phone app that works without an account, but we have not built the full connection yet.
Next we will make enclosed probes that can stay in real beds. We expect dirt, weather and ordinary handling to expose problems that this workbench cannot. We will keep recording those problems alongside the readings.