Artificial intelligence technology is transforming the way that manufacturers operate. Companies in a variety of industries are investing in AI tools to improve efficiency, product development, safety, predictive maintenance and supply chain logistics.

For instance, several Stellantis power train assembly plants are using AI-enabled robot guidance systems. The automaker leverages AI and vision systems to allow robots to adjust their trajectory in real-time to avoid potential conflicts or impacts. The system is delivering improved quality and reducing lead times.

Siemens AG has harnessed the power of AI at its state-of-the-art electronics factory in Erlangen, Germany, where machine learning optimizes test and inspection procedures, significantly increasing first-pass yield and boosting efficiency. AI-enabled robots that pick and place components on fully automated assembly lines have reduced automation costs by 90 percent.

Engineers at the University of Virginia (UVA) are also developing AI-driven systems to transform how factories operate. Using multi-agent reinforcement learning (MARL), they have created a more efficient way to optimize manufacturing systems, improving both speed and quality while reducing waste. By coordinating multiple agents to manage tasks in real time, the system adjusts automatically, learning and improving performance over time.

“AI is revolutionizing the manufacturing industry by offering a wide range of benefits that can significantly improve efficiency, productivity and overall performance,” says Qing “Cindy” Chang, Ph.D., a mechanical engineering professor who is heading up the R&D project.

“AI can analyze vast amounts of data collected from assembly lines, such as individual process times, downtime and over-cycle rates,” explains Chang. “By processing this information, [the technology] can identify bottlenecks, predict potential failures and optimize production schedules.”

However, Chang warns that a big misunderstanding about using AI technology in manufacturing is the belief that it can operate autonomously without human intervention or domain expertise.

“Many assume that implementing AI is a plug-and-play solution,” says Chang. “But, successful applications require high quality data and domain expertise. Customization is also often needed, since manufacturing environments vary significantly.”

According to Chang, one of the key benefits of AI is enhanced efficiency and productivity. Specifically, it enhances three types of factory operations:

• Predictive maintenance. AI algorithms can analyze sensor data from machines to predict potential failures before they occur, allowing for proactive maintenance and minimizing downtime. 
• Process optimization. AI can identify inefficiencies in production processes and suggest improvements, leading to faster production cycles and reduced costs. 
• Automated quality control. AI-powered vision systems can inspect products for defects, ensuring consistent quality and reducing the need for manual inspection.

The technology can also improve plant floor decision making. “AI can analyze vast amounts of data from various sources to provide valuable insights into market trends, customer preferences and future demand forecasts,” explains Chang.

“By integrating diverse information using neural networks, algorithms can train control policies to optimize plant operations,” says Chang. “This includes fine-tuning cycle times, scheduling maintenance, prioritizing resource allocation and optimizing inventory levels to significantly enhance overall performance efficiency.”

Chang and her colleagues are focusing their research on modeling and controlling complex manufacturing systems. “These systems involve hundreds of interconnected machines, robots and human workers, often subject to random disruptions like machine failures and material shortages,” she points out.

“To address this complexity, we're developing neural network models to integrate vast amounts of data and disruption events,” explains Chang. “By employing a multi-agent reinforcement learning control scheme, we aim to optimize system yields, ensuring the production of high quality products.

“Our approach leverages physics-guided MARL to achieve exceptional robustness compared to other machine learning techniques,” claims Chang. “This innovative method enables our system to adapt to dynamic conditions and make real-time decisions, ultimately enhancing overall manufacturing efficiency and productivity."

The UVA research is being conducted in collaboration with General Motors, which provided valuable insights and real-world applications for the AI system. “GM’s involvement ensures the technology addresses the practical challenges of modern manufacturing,” notes Chang.”

“Instead of optimizing individual processes in isolation, our system looks at the big picture—coordinating everything at once,” says Chang. “The result is smarter, faster and more adaptable manufacturing.”

Algorithms were key in making this advancement. They enable the system to account for both the physical constraints of machinery and unpredictable production disruptions.

“By integrating system- and process-level parameters, this system can optimize yields and dynamically adapt to changes, such as machine breakdowns or production adjustments, without human intervention,” says Chang. “It’s a major leap forward in smart manufacturing.”

Chang believes this AI-driven control system could establish new benchmarks for manufacturing efficiency, particularly in complex, multi-stage production environments. She says it sets the stage for smarter, more adaptable production systems, with broad potential applications across various industries. 

“In addition to improving productivity, the system offers significant economic and environmental advantages,” adds Chang. “By reducing waste, minimizing downtime and lowering energy consumption, manufacturers can achieve substantial cost savings while shrinking their environmental footprint. The technology presents a powerful step forward for both industry and sustainability efforts.”

NOTE: This article was researched and written 100 percent by human – AI was not used in content creation.