Skip to main content

AI Becomes Pervasive in Manufacturing, But What Should Users Look For?

By May 22, 2024Uncategorized

AI shows extreme promise for software and product development, manufacturing operations, and engineering and design. Industrial AI, a subset of the broader field of artificial intelligence (AI), refers to the application of AI technologies (including Generative AI) in industrial settings to augment the workforce in pursuit of growth, profitability, more sustainable products and production processes, enhanced customer service, and business outcomes. Industrial AI leverages machine learning, deep learning, neural networks, and other approaches. Some of these techniques have been used for decades to build AI systems using data from various sources within an industrial environment, such as sensors, machinery, industrial engineers, and frontline workers.

Finding the right AI based solutions for the specific and often challenging requirements of manufacturing and industrial applications can be a challenge. AI isn’t a product that you buy, but instead exists within myriad applications, development frameworks, and solutions all designed to accomplish specific functions. Leading industrial organizations align AI initiatives with broader business objectives, ensuring maximum value from technology investments. ARC analysts recently discussed the far-ranging impact of industrial AI with Larsen & Toubro Technology Services (LTTS), who has turned their expertise in software and engineering into a comprehensive suite of AI-based offerings specific to a range of use cases and applications.

Evolving Industrial AI’s Role in Digital Transformation

Industrial AI, a subset of the broader field of artificial intelligence (AI), refers to the application of AI technologies (including Generative AI) in industrial settings to augment the workforce in pursuit of growth, profitability, more sustainable products and production processes, enhanced customer service, and business outcomes. Industrial AI leverages machine learning, deep learning, neural networks, and other approaches. Some of these techniques have been used for decades to build AI systems using data from various sources within an industrial environment, such as sensors, machinery, industrial engineers, and frontline workers.

Focusing AI on Business Results

In industrial organizations, achieving desired business outcomes requires a comprehensive approach that encompasses the essential triad of people, processes, and technology. This framework is not just a buzz phrase, but a proven strategic blueprint that guides organizations toward sustainable success. The convergence of AI with other technologies, such as Internet of Things (IoT) and edge computing, opens new possibilities for distributed and embedded AI systems that can operate at the edge of the network, closer to where data is generated.

Industrial leaders are identifying areas where AI can make a significant impact, such as generative design of sustainable products, production processes, and services, predictive maintenance, supply chain optimization, and quality control.

Use Cases for Industrial AI

Industrial AI solutions and applications are popping up everywhere, but their ability to address specific problems for industrial end users can sometimes be elusive. Many different classes of competitors are entering the industrial AI market, from the large automation suppliers to the hyperscalers to smaller niche suppliers.

The purpose of Industrial AI is multifold. It aims to enhance operational efficiency by automating repetitive tasks, improve accuracy by reducing human error, and enable real-time decision making based on data-driven insights. From generative design of products and production processes to intelligent production operations maintenance and quality control, to energy and supply chain optimization, efficient sales and enhanced customer service, Industrial AI finds its applications across a wide spectrum of industrial operations.

Industrial AI offers several benefits. It can significantly reduce operational costs by optimizing resource usage and improving process efficiency. By enabling predictive maintenance, it can minimize downtime and extend the lifespan of machinery. With its real-time decision making capabilities, it allows for rapid response to changes in market demand or operational conditions, addressing organizational skills gaps, and enhancing agility and competitiveness.

In the realm of product and process design, AI-powered software enables more accurate and efficient design methodologies. Advanced AI algorithms can analyze vast amounts of data to predict optimal design parameters, reducing trial-and-error and speeding up time-to-market. Additionally, AI can assist in simulating and testing product designs under a variety of conditions, ensuring robustness and reliability.

Production operations and maintenance have been significantly enhanced by AI. Predictive maintenance, powered by Machine Learning algorithms, can anticipate equipment failures before they occur, reducing downtime and maintenance costs. AI can also optimize factory operations by streamlining workflows, improving resource allocation, and enhancing quality control.

The logistics of supply chain management are becoming increasingly complex in our globalized economy. AI comes into play here by adding predictive analytics for demand forecasting and inventory management to real-time visibility. It can also optimize routing and scheduling for transportation, leading to reduced costs and improved customer service.

The LTTS Approach to Industrial AI

As a large engineering service provider with considerable strength in software development, LTTS is uniquely positioned to provide value-added solutions to the world of industrial AI. LTTS’ knowledge of specific processes, plants, and facilities can add significant value to solutions. LTTS views AI as one set of enabling technologies, along with IIoT, digital twins, simulation, and other technologies, which can provide value both in the engineering realm, the supply chain, and the domain of manufacturing.

A Business-Driven Approach to Industrial AI

LTTS aligns its service offerings with the business needs of its customers, rather than focusing on technology alone. By understanding the pain points and cost drivers of different industries, LTTS can provide tailored AI solutions that address specific problems and deliver desired outcomes. LTTS has demonstrated the value of AI in improving operational efficiency, reducing costs, enhancing quality, and increasing customer satisfaction.

Industrial AI as an Application Layer

LTTS leverages its expertise in engineering and software development to create AI-powered applications that turn data into knowledge. By applying machine learning, deep learning, neural networks, and other AI techniques, LTTS can build intelligent systems that can analyze data, generate insights, and automate tasks.

Industrial AI as a Partner Ecosystem

LTTS collaborates with technology providers such as Qualcomm, Nvidia, Google, AWS, and Intel to deliver innovative and customized AI solutions. LTTS also works with customers to co-create and co-innovate AI applications that suit their specific needs and goals.

Augmenting Worker Capabilities with Industrial AI

LTTS uses AI to enhance the skills and productivity of the workforce, rather than replacing them. By providing AI tools and training to the frontline workers, LTTS can empower them to make better decisions, improve their efficiency, and reduce human errors. LTTS has plans for training 2,000 Engineers in Gen-AI using technology from Nvidia.

AiKno Metadata Extraction

One common factor amongst all industrial companies is the massive amount of paperwork & manually entered data from the past. The LTTS Cognitive Meta Data Extraction Module automates the tedious task of digitalizing physical data, thereby improving efficiency by 92-95 percent and reducing the time by 85 percent.

What makes the solution unique is its ability to extract metadata from complex engineering documents such as 2D drawings, legacy documents, and scanned images. The OCR is trained in engineering symbols such as GD&T symbols which are heavily used in the industry. The continuous self-learning system can do auto corrections and drive semantics-based rules on human feedback without any need for re-engineering. As a case in point, AiKno helped optimize the PO/RO processing for a transportation major. The manual digitalization of 1,000 documents would take the team 233 hours. With AiKno, this time was reduced by 60-80 percent to just 42 hours.

Predictive Analytics

The LTTS AiKno predictive analysis framework offers real-time insights into equipment health and identifies anomalies or failures well before they actually occur. Using built-in AI/ML models, service requests are automatically initiated, or machines run self-diagnosis programs to resolve issues. The LTTS AI framework can automatically preprocess and run different ML algorithms and evaluate the models on the basis of key metrics. It can automatically choose the best possible model and lower the manual effort of creating the models.

Using AI to Create an Asset Health Framework for a Global Food and Beverage Company

LTTS is actively implementing AI-based projects today with measurable results. The company recently implemented a predictive asset health framework for a major global beverage company to address primary asset performance management related issues such as change management and a lack of consistency in deploying solutions across multiple plants. Unplanned downtime is a significant source of loss for this industry, particularly in areas such as bottling and filling lines.

The end user wanted to transition from a preventive maintenance philosophy to a predictive one to improve uptime and OEE. The user wanted a solution that could be deployed across multiple plants in the US. LTTS provided an AI-based predictive maintenance solution across 14 critical assets in one plant, in areas such as the blow molder, fillers and cappers, labelers, packers, and ammonia chillers. The solution featured over 200 sensors with 34 edge gateways, providing real-time health monitoring of over 100 asset components. LTTS deployed analytics at the edge, enabling early fault detection combined with reduced latency and data congestion. The edge analytics were deployed using pre-built models along with custom machine learning for greater accuracy. The impact on the business was substantial; availability of the production lines improved significantly, with over 17 hours of unplanned downtime avoided in an 8-month period, with potential savings of around $300,000.

Conclusions

Understanding AI’s potential implications and benefits, and the ability to effectively wield it, is crucial for industrial organizations. Artificial Intelligence can be a daunting task, but with the right strategies and approaches manufacturers can increase their odds of successfully implementing AI into their operations. Software providers have a lot of technical expertise, but in many cases, they focus on technology but have comparatively smaller services and solutions businesses that can both implement and maintain a solution, which is really the key to obtaining business value. Finding solution providers with the right experience implementing these solutions in specific use cases can be a challenge.

LTTS is unique among many of the companies providing AI solutions for industrial applications because of both their depth in engineering expertise and software development. The company’s engineering expertise gives unique capabilities in accessing engineering data, and LTTS has considerable expertise in digitizing data and turning it into useful information that can be ingested by AI models to create an effective solution.

The post AI Becomes Pervasive in Manufacturing, But What Should Users Look For? appeared first on Logistics Viewpoints.

Leave a Reply