Maximizing Efficiency in Data and Power for IoT Camera Devices
Author: Daniel Chen
Abstract
With the rise of homemade, small-batch IoT devices, creating functional systems has become more accessible thanks to open-source libraries and pre-made hardware. However, an efficient IoT camera system must consider two key factors: data transfer and power consumption. This study explores the best solutions to manage data transfer and power consumption for IoT camera systems delivering Computer Vision.
Research Context
AI cameras and computer vision systems typically run in environments with stable power and internet connections, like CCTV systems with existing wiring or standalone cameras with stable Wi-Fi and USB connections. Current solutions often involve brute-force methods, such as adding larger batteries or 5G connectivity. This research aims to find more efficient, smaller-scale solutions focusing on data transfer rates and protocols.
Researching Method
Selecting the best data transfer method from numerous available options is crucial. Using an STM32 platform with a standard setup (STM32H7 Microprocessor and OV5640 camera), I will test various methods through trial and error. The goal is to identify the most efficient method for data transfer while minimizing power consumption.
Research Questions
What is the best way to transfer low latency large data packets?
What are the physical limitations of current technologies?
What is the smallest physical size needed for this system?
What is the financial cost at small and large scales?
Central Aim
To find the most efficient method for transferring low latency large data packets for IoT camera systems, considering power consumption, data throughput, and overall cost.
Results
