“TinyML model running Wio Terminal can identify if animal or human is approaching even in the dark, sends data to Helium LoRa network to AWS.
In the previous project, I used Wifi connection to send tinyML inference data to AWS IoT. As you can imagine, this device is meant for camping, away from your home wifi network, LoRa makes more sense instead of wifi connectivity. In this project, I am using Helium network which is the most popular choice these days due to broader network coverage and it’s growing rapidly.
I am going to use same Edge Impulse model. If you are interested in tinyML part, please checkout my previous project. Here, I am going to explain how to hook up with helium and integrate helium with AWS.
Required hardwares needed for this project are listed above. Connect the LoRa antenna with chassis, connect thermal camera to I2C port of battery chassis, connect servo motor to digital pin 0 of battery chassis. Then stack battery chassis on top of LoRa chassis and wio terminal on top of battery chassis.
Then upload the sketch WIO_Camper_Inference_lora.ino to Wio-Terminal.
LoRaWan
Before you go further, you need to understand some basics and may need to modify above code accordingly. LoRaWAN is a low-power, wide area networking protocol built on top of theLoRa radio modulation technique. It wirelessly connects devices to the internet and manages communication between end-node devices and network gateways. Each country has specific frequency bands and data rates (DR). Depending on DR, max payload size varies. In my program I am using US915 with DR 0 which gives me maximum payload size of 11 bytes. Ideally DR 2 works with helium but I could not get it working with wio chassis, so stuck with DR0 but it’s alright as my payload size is 10 bytes.”