Using ClusterDuck Protocol to Monitor Vacuum and Temperature in Maple Syrup Production

Jeff Wiles
ClusterDuck Protocol
5 min readMar 28


Photo by Kathyryn Tripp on Unsplash


At our outdoor camp in Ohio, we utilize over 200 acres of maple trees to create delicious maple syrup. Our process involves tapping each tree with 1–2 taps and connecting them to lateral 5/16" tubing. These lines are then connected to 1 of 7 main lines, made of 1" plastic tubing, which transport the sap downhill to sap collection tanks. To maximize production, we have implemented a vacuum system that runs throughout the entire system. This innovative technology has had a significant impact on our syrup production, allowing us to create more syrup than ever before.

Our camp contains a total of over 900 taps, connecting 700 trees in total. On a good day over 2000 gallons of sap can be collected. It takes 60–70 gallons of sap to make 1 gallon of syrup. Last year we collected 24,000 gallons of sap and produced close to 300 gallons of finished syrup!

A main line racing across the forest.

As you can imagine, producing maple syrup at this scale is a highly labor-intensive process, and we rely entirely on volunteers to make it happen. However, by automating certain aspects of the process, we are able to significantly increase efficiency and streamline our operations. By leveraging technology to our advantage, we can produce maple syrup more efficiently and ensure that we are able to provide a high-quality product to our customers.

Vacuum and Temperature Monitoring

With all the tubing and the connectors, there are many places where the tubing can leak and loose vacuum. If it does leak, it would be very hard to pinpoint where it is in the 200 acre of land. For a solution, I created sensors built with Arduino micro controllers to monitor vacuum and temperature. If there is any abnormality in the readings, then we would know exactly where the leak is and fix it right away.

Heltec board / solar charge controller / vacuum sensor

These devices are placed at the ends of each of the main sap lines. If it is deployed at the end of each main sap line, how do we get the readings?

The first option was utilizing Wi-Fi. We can use Wi-Fi to send sensor readings to the cloud. However it was not the optimal solution. One major problem was power. Wi-Fi consumes a lot of power and there is the lack of electricity in forests. Previously, I used a 12 volt car battery connected to a Wi-Fi access point to provide Wi-Fi. Sensors would connect to Wi-Fi and publish sensor readings. Another problem was the distance. The sensors are 300–400 yards from the closest relay/gateway device. Wi-Fi was not supportive of long range communications. The last problem was unreliability of Wi-Fi. Obstructions (trees, hills) and weather made Wi-Fi access points unreliable at times.

This is where the ClusterDuck Protocol (CDP) comes in. CDP is an open-source firmware that leverages LoRa to wirelessly communicate with internet-of-things devices called Ducks. CDP allows for greater distances without the need to adding Wi-Fi access points out in the woods. And best of all, it is a low power wireless ad hoc network (LPWAN). This simplifies the process and, in my experience, is much more stable than Wi-Fi.

We integrate the CDP into the sensors so that all the data from the sensors relay through the trees back to our maple barn. Leveraging CDP allowed us to monitor the maple syrup production in real-time.

Sensor connected to mainline.

Design of MapleDuck Network (2023)


  • Heltec WiFi LoRa 32 V2 boards
  • Onewire temperature sensor
  • Adafruit MPRLS Ported Pressure Sensor
  • 90X60mm Micro Mini Solar Panel Cells
  • 18650 battery


  • gateway LoRa + Wi-Fi
  • TTGO T-Beam board — Connected to power

Raspberry Pi

  • Apache web server
  • Mosquito MQTT server

How Does The Network Deliver Data

The sensors take readings of vacuum and temperature from the environment every 10 minutes. Then they broadcast these readings to the to nearby devices and then goes to sleep (to preserve battery) until it is time to read the sensors again or it receives a message to relay.

The messages are relayed from device to device until it reaches the PapaDuck (the gateway). When the PapaDuck sees the LoRa message, it decodes the CDP message and publishes a message over Wi-Fi to MQTT server.

In the maple barn I have a Raspberry Pi running Mosquito, and Apache. As the readings are published to the server, they are displayed on a web page. With the current setup 90X60mm Micro Mini Solar Panel and 18650 a battery, several of the sensors have been running 5 weeks with no issues.

Dashboard showing vacuum of main lines.

Tank Level and Temperature Monitoring

Along with the Vacuum/Temperature sensors, I have built Tank Level Sensors. These use a Maxbotix Weather-Resistant Ultrasonic Rangefinder — MB7092 XL-MaxSonar-WRMA1 to measure the distance from the top of a tank to where the sap is located. We have 6 different tanks varying in sizes from 150–1500 gallons. Most of these are retired bulk milk tanks, horizontal cylinders. I convert the volume into gallons of sap. These sensors are using WeMos D1 WiFi Arduino Development Boards. Readings are taken every 30 seconds a sent to the MQTT sever. These are close to power and Wi-Fi, so I have not converted these to LoRa

Dashboard showing vacuum of main lines.

At the moment, I am working with the ClusterDuck Protocol open source community to provide some feedback and improve the network. In addition, I am in the process of finalizing the microprocessing board. I used Heltecs but maybe CubeCell might be a better option. Also Onewire temperature sensor integrated with CDP network. The end goal will be automating the entire process to significantly increase efficiency and streamline our operations.