42Technology helped Titan Class, an Australian Internet of Things company, develop a faster and easier way to extend its smart agricultural monitoring system to some of the most remote farms in the country without farmers needing to install expensive infrastructure such as gateway antennas.
new Solutions that support NB-IoT Use existing off-the-shelf agricultural sensors, combined with the Rust programming language application on the Nordic Semiconductor nRF9160 SiP device, to transmit small LoRaWAN-style sensor data packets to the network. As a result, Titan Class reduces the transmission cost of sending real-time data through a cellular network by 100:1, thereby reducing network costs and enabling the system to handle more sensors than ever before.
This method has been tested with a soil moisture probe, but it can be used for any farm monitoring or driving task, including monitoring the weather, the water tank level, the status of a fuel tank, or an electric fence.
Although Australia’s NB-IoT network has a high coverage rate, one of the challenges is that the sensors on site may be as long as 20 kilometers from the nearest base station. Since it is almost impossible to maintain full-duplex cellular communication over these distances, NB-IoT can be used for “fire and forget”. This is in sharp contrast to the TLS-based MQTT settings used by most agricultural sensors: the full-duplex nature means that they must use LTE-M networks, which leads to unnecessary high network costs because the MQTT protocol has relatively high overhead. Each data packet used for connection establishment and transmission.
Christopher Hunt, chief technology officer and co-founder of Titan Class, sells its smart agricultural monitoring system under the Farmify brand. He said:
“Farmers told us that they usually pay about US$50 per month in network fees for each device, and our preliminary experiments have shown that we can send LoRaWAN packets over UDP to reduce this cost to about US$0.58. Our method also More robust and reliable, because we get higher coverage and penetration rate from NB-IoT compared to LTE-M network.”
The role of 42T in this project includes improving the support for nRF9160 in the existing Embedded Rust’nrf-hal’ project, and updating its own Rust language open source interface to the Nordic modem to use the latest version of the Nordic library.
“The Rust programming language aims to provide a more efficient way to build high-performance, reliable software. It is now widely used in back-end systems, and all major technology companies have invested heavily and reported that its launch has been a huge success,” Said Jonathan Pallant, senior embedded system developer at 42 Technology.
“This kind of support has also begun to appear in the field of embedded Rust. The combination of high-performance code and attention to safety and reliability means that it is very suitable for applications such as this NB-IoT-based agricultural monitoring system.”
