Sensor data fusion for optimal robotic navigation using regression based on an IOT system

• Published in: Measurement. Sensors Vol. 24; p. 100598
• Main Authors: Aroulanandam, Vijay Vasanth, Satyam, Sherubha, P, Lalitha, K, Hymavathi, J, Thiagarajan, R
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.12.2022; Elsevier
• Subjects: And regression; Estimation error; IoT; Machine-to-machine; Navigation; Robust; Sensors; Estimation error; Navigation; And regression; Robust; Machine-to-machine; Sensors; IoT
• ISSN: 2665-9174; 2665-9174
• DOI: 10.1016/j.measen.2022.100598

Engineers can now build compact gadgets with potent processing power and network capabilities thanks to the semiconductor industry's rapid growth. The Internet of Things is one of the greatest subjects in the communication world right now, both professionally and academically. The technical idea behind the Internet of Things is to provide these various physical items the ability to use sensors to perceive information and transfer that data to a server. The goal of the industrial Internet of Things (IIoT) is to integrate multiple technologies to enhance business services across various industries. It alludes to how machine-to-machine connections operate. Every industrial Internet of Things (IIoT) service domain has unique communication needs that are assessed differently in terms of efficiency, dependability, and quality of service (QoS). For example, there are many surveillance applications for an Internet of Things-based robot with AI. To get around the problems with current robotic navigation, we suggest an Internet of Things (IoT) platform with an intelligent navigational robot that uses machine learning. It was designed with its Internet of Things (IoT) server, and it can do line tracing, gather contextual data from the sensors, and identify the aberrant condition in an environment before navigating appropriately. To allow contextually aware IoT platform services and to validate the learning accuracy of the provided CNN model using a dataset of images collected from the robot, we developed the Support-Vector Regression with Improved Adaptive Lion Optimization (IALOSVR) Algorithm model. According to experimental findings, learning accuracy exceeded 0.98, indicating that we had improved learning in picture context identification. The real-time evaluation of this paper's method demonstrates its ability to estimate navigation accurately, robustly, and with the estimation error, high estimation accuracy is steady within two percentage.