IOT TELEGRAM BOT NETWORK DEPLOYMENT AND MEASUREMENTS
Loading...
Date
2024-04-19
Authors
Bolatov, Adlet
Salym, Adil
Journal Title
Journal ISSN
Volume Title
Publisher
Nazarbayev University School of Engineering and Digital Sciences
Abstract
This project was done to resolve connectivity inconsistencies encountered while
transmitting data from devices with temperature and luminosity sensors to a Telegram bot due to instability in WiFi connection. Consequently, Long Range (LoRa) technology was implemented in this project due to its relevance for low-power and wide-area radio communication. The project lifetime lasted from September 2023 to April 2024, in a total of 7 months together with a 1 month break during December and January. Initially esp32 and Pycom LoPy4 were used as devices, however later we switched to both of them to be esp32. Sender device reads data from temperature and luminosity sensors and sends them to the receiver. Aside from that, the receiver device uses OpenWeatherMap API to get the outside temperature in Astana with a given longitude and latitude. Additional functionality has been added to this project like connection to OpenAI API and machine learning implementation. For the OpenAI API part, the receiver device has been connected to the API to access ChatGPT queries and get answers for questions
from it. The ML model implementation uses data gathered from the telegram bot for the past 10 months, which is stored in .csv format. At the end of the day the receiver device connects to OpenWeatherMap API and gets predictions for morning and afternoon temperatures outside of university, using these temperatures and ML model, it posts predicted temperatures for atrium and outside to telegram bot. In the end, multiple tests were conducted with LoRa and other functionality, so that one user uses a sender device and another a receiver device. The connection was tested from different spots, and the main spot for receiver was C4 block in Nazarbayev University, and main spot for sender was green spot in the atrium of university. The connection resulted in a range of -94 to -97 RSSI value which is acceptable for LoRa connection, solving the main problem of the project.
Description
Keywords
LoRa, Internet of Things, ChatGPT, Machine Learning, esp32, Telegram bot, Type of access: Restricted
Citation
Bolatov, A., & Salym, A. (2024). IoT telegram bot network deployment and measurements. Nazarbayev University School of Engineering and Digital Sciences