Peat (commonly called turf in Ireland) has been a source of fuel for heating and cooking in Ireland and other parts of Northern Europe for centuries (if not millennia). The traditional method of peat harvesting involves cutting turf from a peat bog and then setting it out to dry.
A long-time colleague, Martin Bradley who teaches Electronic Engineering at Letterkenny Institute of Technology (LYIT) came up with an interesting method using PSoC 4 and resistive sensing to measure the moisture content of cut turf to know when it is dry enough to turn, foot or collect.
From Martin’s description of the project:
The turf was cut in Ballinlough, Drumfries, Clonmany, Co Donegal, Ireland, moss (hill) on Friday 19th April by Kevin Doherty, a neighbour of mine and I connected a PSoC4 monitoring device (in deep sleep mostly, wakes up every 10 minutes) with wireless capability (Sigfox) to it at noon on Monday 21st. The aim is to get data that would be a good estimate for when the turf should be turned or then “footed.”
Interested parties, such as farmers could be notified by phone when action is required. It is meant to be a bit of craic (fun) but also may provide useful information. It is certainly starting loads of discussions. e.g. how much will the turf be set back if there is a very wet day? Will it be the same value as a week previous or more?
Here is a picture of how the experiment was set up as well as a graph from MATLAB’s “thingspeak” cloud-based tool. The PSoC 4 is the red PCB in the plastic box (to protect it from the elements).
The graph drops with resistance rise. This drop should indicate turf drying. Results show that so far resistance is rising indicating dropping moisture levels in the turf although there are probably other things happening that may affect resistance of turf e.g. morning dew, temperature, chemical reactions etc. A second test is ongoing with the probes inserted in both ends. This reduced the effect that surface water (due to rain) on the turf had.
It was noticed that the running average is certainly a good indicator of the turf drying. When the value goes below a certain threshold value, free notifications are sent to mobile devices (smartphones or tablets) to inform users. This was achieved by a combination of the Sigfox network, thingspeak.com and ifttt.com.
A similar setup that uses Lidar is being used to measure the water level in rivers with notifications (early flood warning etc).
If you would like to see current live data click here
This experiment also raised the interest of Holger Wech (Sr. Staff Applications Engineer at Cypress Langen, Germany), who created a similar project to automatically water flowers some time ago. Here is the dialog. The questions are Holger’s, the answers are Martin’s.
HW: Have you already measured the power consumption of your system?
MB: PSoC4, TD1207R Sigfox module, 1 resistor plus 2 x AA batteries
4.5uA for complete system in deep sleep mode
50mA when transmitting for 5 seconds
Wakes up every 10 minutes, measures resistance and transmits and then goes
back to deep sleep
HW: Why did you choose SigFox?
MB: I have used Sigfox for numerous projects. low power was one consideration
HW: Have you thought about your electrodes to prevent from any kind of oxidation?
MB: I did but the lifetime of the test will hopefully rule this out as a problem
HW: I remember, it’s best to use AC measurement instead of just DC voltage.
MB: I used DC as it was easier to set up for low power etc.
This system does not need to be extremely accurate.
Update: The Donegal Daily has recently published an article regarding Martin's project
I hope you enjoyed learning about Mr. Bradley’s unique project.
Until next time C. U. A. round!