Download PDFOpen PDF in browser

An Industrial Self-Learning Robotic System Platform Solution for Smart Factories Industrial Applications Using Machine and Deep Learning

EasyChair Preprint no. 6104

6 pagesDate: July 16, 2021

Abstract

Artificial Intelligence, Robotics and Industrial automation are developing rapidly. As they are progressing from simple reasoning to improving and strengthening human capacity, increasing productivity, and human-like cognitive abilities. This confirms the urgent requirement to adopt smart factories and improve manufacturing facilities in order to meet the massive challenges facing the UK industry moving into digital and green technology. Smart Factory is an important platform for recent industrial development, and industrial robot systems are an essential measure of the configuration and recombinant of its cells. As it can reduce working time and give less human error when performing high-precision processes. It is one of the excellent solutions to achieving dynamic, programmable and fixed industrial automation requirements. As its flexible software platform can usually be modified easily to meet the specific requirements and current market demand of end-users. This makes it a cost-effective, portable and low-maintenance solution for many industrial applications where manufacturing is required: wide range of products and parts, specific classes of product changes, and fixed order of processing or assembly. Moreover, there are some significant gaps between the offering of the industrial robot systems and the expectations of industrialists. The main purpose of this research is to develop a self-learning robotic system platform solution using Imitation and Deep Learning for smart factories industrial application. The Aim is to improve the ability of configuring robotic systems to make more flexible decisions and cost-effective manufacturing environments.

Keyphrases: Automation, degree of freedom, Imitation Learning, industrial robots, Smart Factories

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@Booklet{EasyChair:6104,
  author = {Yagna Jadeja and Mahmoud Shafik and Paul Wood and Leonardo Stella},
  title = {An Industrial Self-Learning Robotic System Platform Solution for Smart Factories Industrial Applications Using Machine and Deep Learning},
  howpublished = {EasyChair Preprint no. 6104},

  year = {EasyChair, 2021}}
Download PDFOpen PDF in browser