Contemporary Trends and Challenges and Advances, in the Manufacturing Industry, with a special focus on applications of Artificial Intelligence and Deep Learning

Abstract

In the last few years’ artificial intelligence (AI), has begun to make its appearance in our everyday life. Even though it is still in its early stage of development, AI has proved beyond human intelligence. DeepMind’s AlphaGo is an illustration of how the AI could provide amazing benefits, particularly in industries such as manufacturing. At the moment there are attempts to connect AI technology with precision engineering and manufacturing in order to change classical production methods.

This research paper focuses on some notable milestones that have already been attained in the realization of AI for manufacturing and how it will change the face of any manufacturing facility. There are several challenges in the AI manufacturing application; these include data acquisition and management, human resources, infrastructure, security risks associated with trust issues as well implantation of hurdles. For instance, the collection of data required to train AI models can be challenging for rare events or expensive in large datasets that require labeling.

The introduction of AI models into industrial control systems can also pose risks to the security, and some players in industry may be reluctant to use AI because they don’t trust it or understand what is going on. However, these hindrances do not deter AI from becoming an effective solution for predictive maintenance and quality assurance in the sector of manufacturing. Therefore, one should ponder over each manufacturing case and its needs before deciding if or how to adopt AI.

The aim of this research paper is to analyze the current progress, problems and prospects in AI-ML across manufacturing entities. Its aim is to enhance knowledge of accessible technologies, support decision-making in choosing appropriate AI/ML technologies and determine where further research needs are possible centered on latest developments. Initial findings indicate that the combination of AI/ML technologies with advanced data collection capabilities from manufacturing networks can produce massive cost and efficiency gains.

Though the accurate representation of complex phenomenon in manufacturing is problematic, AI can revolutionize this industry. Other areas where AI is intensively studied include medical image analysis, bioinformatics, recommendation systems and finance. Many notable AI products such as Amazon’s Alexa, IBM Watson and DeepMind AlphaGo have already integrated into our daily use. To address limitations such as interpretability and degraded performance with insufficient data, several sub-branches of deep learning are currently researched namely; Physics - Informed Deep Learning (PIDL), Explainable AI(XAI), Domain Adaptation, (DA) Active Leaning (AL), Multi Task Learning MTL , Graph Neural Network GNN.

Convergence of AI with other engineering industries have a potential issue that should not be ignored. The aim is to enable an effective use of AI by the precision engineering and manufacturing community for future-oriented manufacture.