Abstract
Digital transformation has become very crucial in global market. Digital technologies have revolutionized the way industries operate, introducing the concept of “Industry 4.0” or the “smart factory”. Digital technologies have substantially transformed the business and society, bringing fundamental changes through the new emerging approaches of the circular and sharing economy. Artificial intelligence (AI) is now become an essential part in the manufacturing industry to enhance the performance and boost the demand and productivity in the manufacturing firms. However, there are some barriers and challenges face by the manufacturing firms in implementing the AI due to the infancy stage of AI in Malaysia such as lack of talent, lack of incentive and innovation. Therefore, the purpose of this study is to identify the barriers and challenges in implementing AI. In this paper, managers from the manufacturing companies were selected as respondents based on Federal Manufacturing Malaysia Directory. The questionnaires are distributed to the respondents by using the online survey. A total of 93 questionnaires was collected with feedback rate, 23.3%. Descriptive analysis is used to identify the barriers and challenges. The highest level for barriers of AI is lack of talent. The highest level for challenges is no experts in company. This research provide input for manufacturing companies to improve the barriers and challenges for future improvement in implementing AI.
About This Research
Publisher: Journal of Advanced Research in Applied Sciences and Engineering Technology Year: 2022 Type: Applied Research Citations: 24
Relevance
Industries: Manufacturing Pillars: AI Readiness & Strategy Use Cases: Cybersecurity & Threat Detection Regions: Malaysia, Southeast Asia
Legacy Infrastructure and Technical Debt
Manufacturing environments present unique infrastructure challenges that distinguish them from service-sector digitalization efforts. Proprietary programmable logic controller architectures, heterogeneous sensor protocols, and decades-old supervisory control systems create integration bottlenecks that contemporary AI platforms are ill-equipped to navigate without substantial middleware development. The research quantifies the hidden costs of technical debt remediation, demonstrating that infrastructure modernization prerequisites typically consume forty to sixty percent of total transformation budgets before any AI capability becomes operational.
Shopfloor Culture and Knowledge Transfer
Experienced machine operators possess invaluable tacit knowledge about equipment behaviour, material properties, and process optimization heuristics that formal documentation rarely captures. AI systems designed without incorporating this embodied expertise produce recommendations that operators instinctively distrust, creating adoption resistance rooted in legitimate concerns about safety and product quality rather than technophobia. Successful transformation programmes invest heavily in participatory design workshops where shopfloor personnel contribute their experiential knowledge to model training and validation processes.
Supply Chain Ecosystem Readiness
Individual manufacturer AI adoption cannot proceed independently of supply chain ecosystem digitalization maturity. Predictive maintenance algorithms require component supplier quality data, demand forecasting models depend on downstream distributor inventory signals, and logistics optimization systems need carrier telematics integration. The research identifies supply chain ecosystem readiness as a frequently overlooked constraint, with enterprises discovering that their own internal transformation pace is ultimately bounded by the digital maturity of their weakest supply chain partners.