At the PepsiCo shop in Coventry, where popular snacks like Nacho, Walkers, and Wotsits are manufactured, products move swiftly along a conveyor belt. The hum of advanced technology fills the air, creating a nearly overwhelming ambiance.
Amidst the bustling factory environment, it’s not just the human ears that are attentive for signs of equipment malfunction. Specialized cameras, trained to detect the sounds indicative of worn-out machinery that could disrupt production lines, are also actively monitoring for any technological glitches.
After a successful trial in the United States, PepsiCo is now implementing these cutting-edge sensors, developed by the tech firm Augury and powered by artificial intelligence (AI), across its manufacturing facilities.
This initiative is part of a broader trend where businesses are exploring how AI can enhance efficiency, reduce waste, and accelerate product development.
- Explanation of AI Functionality and Definition
- A Simple Guide to Enhance Your Understanding of AI
Delving into the Data
AI is playing a pivotal role in the modern wave of manufacturing, spanning from initial design to product distribution.
Thanks to its ability to analyze vast amounts of data, manufacturers can now anticipate and mitigate potential disruptions proactively.
The financial stakes are high, with companies risking significant losses due to even brief periods of factory downtime. Lengthy delays can result in an inability to meet consumer demands, especially during peak seasons like holidays or events such as Black Friday.
As a response, factory floors are witnessing the integration of real-time monitoring tools that can analyze processes, alert operators to imminent issues, and leverage historical data to recommend operational adjustments.
To detect anomalies such as wear on conveyor belts and joints by analyzing equipment vibrations, the sensors utilized in PepsiCo’s facilities have been trained on extensive audio datasets.
According to Saar Yoskovitz, the CEO of Augury, “We have analyzed and monitored over 300 million hours of equipment operation, enabling us to develop algorithms that can pinpoint specific patterns indicative of various malfunctions.”
This technology empowers maintenance personnel to schedule proactive servicing and prevent reactive responses to machine failures. Moreover, it provides insights into overall equipment health, predicting potential future failures.
Employing AI-driven sensors also aids in reducing energy consumption across PepsiCo’s operations.
By ensuring optimal operational efficiency, businesses can minimize unnecessary energy usage, as highlighted by Mr. Yoskovitz.
Another application of AI in various industries involves computer vision technology, which enables the identification of product defects on a large scale. This entails training devices to recognize materials in images and videos.
In fast-paced manufacturing environments with products swiftly moving through sorting systems and conveyor belts, minute defects can easily go unnoticed. This is particularly critical for intricate components like circuit boards and computer chip wafers with complex designs. AI-powered cameras can now detect errors that may have been overlooked by human inspection, using algorithms tailored to identify specific surface irregularities.
- Comprehensive Glossary for Understanding Artificial Intelligence (AI)
Enhancing Operational Efficiency
The integration of AI to enhance operational efficiency in manufacturing processes, including predictive maintenance and quality control, is now considered a standard application of this technology, according to Alexandra Brintrup, a professor of modern manufacturing at the Institute for Manufacturing, Cambridge University.
She envisions that the most promising applications of AI in manufacturing will involve novel capabilities that were previously unattainable, such as optimizing truck sharing in logistics networks, enhancing supply chain transparency, and facilitating resource sharing among manufacturers.
Due to the intricate nature of supply chain networks and the reluctance of some stakeholders to disclose sourcing information, many aspects of production have traditionally been opaque.
However, AI can be leveraged to analyze and predict supplier identities and locations, enabling businesses to navigate obstacles more effectively and empowering consumers to make informed decisions about the origins and components of products.
Prof. Brintrup leads the Supply Chain AI Lab at the Institute for Manufacturing, which has developed predictive algorithms to trace the usage of substances like palm oil, even when labeled under different names in products.
Research from the lab reveals that there are approximately 200 distinct names for palm oil in the US, potentially misleading environmentally conscious consumers.
According to Prof. Brintrup, “Enhancing visibility in the supply chain and providing this information to consumers will become increasingly crucial in a world that is highly attuned to the environmental and social impact of manufacturing.”
Impact on the Workforce
One of the challenges posed by the increasing adoption of AI tools in manufacturing and supply chain operations is the potential impact on the workforce.
Some companies are exploring how AI can enhance workplace safety for production line workers by utilizing machine learning and machine vision techniques to monitor camera feeds for potential hazards or accidents.
To prevent injuries or strain from tasks involving heavy lifting, AI-powered devices like exoskeletons are being deployed in various UK facilities.
Watch the video demonstration by enabling JavaScript on your computer.
David Schwartz, the Global Vice President of PepsiCo Labs, views Augury’s sensors and AI technologies as more than just tools to future-proof factories; he sees them as mechanisms to deliver greater value to employees and customers.
In his words, “It’s about enhancing work processes to increase efficiency in meeting the needs of our workforce and customers, while also preparing for future requirements on a daily basis.”
