Cellular based IoT using oneM2M and Zephyr RTOS

Overview

Our sponsor asked us for sample oneM2M sensor and actuator applications run on a Thingy:91 using the Zephyr RTOS codebase. To achieve these goals and contribute these oneM2M samples, we created a cellular smart IoT irrigation system that uses sensor values to determine whether the valve (actuator) should be on or off, and a data processor to determine valve state from sensor values. We also created a dashboard to manage and display these oneM2M connected devices – seeing sensor values, editing settings, and manually controlling valves. These are all connected to one central oneM2M ACME server that all aspects of the project tree communicate through.

Objectives

The initial goal of our project was to incorporate oneM2M architecture into design of a sample sensor and actuator program for the cellular Nordic Thingy:91 chip. To provide a more specific application, we wanted to create a demo scenario with an irrigation system that showcases the functionality and usability of oneM2M.

Approach

– Discussed potential applications of oneM2M sensor and actuator considering the cellular capabilities of the Nordic Thingy:91 – decided to proceed with a smart irrigation system

– Discussed with somebody in the irrigation industry to find out the problems with the modern irrigation system

-Sensors are lost in the field after running out of power, no location or GPS

-Many irrigation systems lack management for the common user, and you often bring in a specialist to help with modifying settings or fixing
problems

-Project Deliverables

-a sensor to retrieve values like soil moisture, rainfall, temperature, humidity, GPS and more to send to ACME CSE (server)

-an actuator to control the valve based on oneM2M notifications

-a dashboard to provide easy access to sensor info as well as modifiable settings, linking sensors and actuators, and direct valve control

-a data processing calculator to determine what the valve state would be based off of sensors linked to the respective actuator

-The team additionally submitted the project for the international OneM2M Hackathon and received second place for use of cellular and advanced
features with oneM2M

Outcomes

This project provided sample programs for oneM2M sensor and actuator implementations using cellular based devices running on Zephyr RTOS. Through this contribution, others can use the code for other applications with minimal refactoring, as the codebase for communicating with a central CSE via oneM2M is already implemented in our code. This in particular is our main contribution, as the oneM2M standard has not been implemented widely on cellular IoT devices. Our dashboard contributes an effective way to display such information from the oneM2M CSE, and our data processing application shows how these values can be processed on the cloud to decrease strain/processing requirements on physical IoT hardware.