Cloud Hosting

Introduction

Technologies such as machine learning (ML), artificial intelligence (AI), and Generative AI (GenAI) unlock a new era of efficient and sustainable manufacturing while empowering the workforce. Areas where AI can be applied in manufacturing include predictive maintenance, defect detection, supply chain visibility, demand forecasting, product design, and many more. Benefits include improving uptime and safety, reducing waste and costs, improving operational efficiency, enhancing products and customer experience, and faster time to market. Many manufacturers have started adopting AI. Georgia-Pacific uses computer vision to reduce paper tears, improving quality and increasing profits by millions of dollars. Baxter was able to prevent 500 hours of downtime in just one facility with AI-powered predictive maintenance.

However, many companies struggle (per recent World Economic Forum study) to fully leverage AI due to weak foundations in organization and technology. Reasons include lack of skills, resistance to change, lack of quality data, and challenges in technology integration. AI projects often get stuck at a pilot stage and do not scale for production use. Successfully leveraging AI and Gen AI technologies requires a holistic approach across cultural and organizational aspects, in addition to technical expertise. This blog explores how an AI Center of Excellence (AI CoE) provides a comprehensive approach to accelerate modernization through AI and Gen AI adoption.

The manufacturing industry faces unique challenges for AI adoption as it requires merging the traditional physical world (Operational Technology, or OT) and the digital world (Information Technology, or IT). Challenges include cultural norms, organizational structures, and technical constraints.

Factory personnel deal with mission critical OT systems. They prioritize uptime and safety and perceive change as risky. Cybersecurity was not a high priority, as systems were isolated from the open internet. Traditional factory operators rely on their experience gained through years of making operational decisions. Understanding how AI systems arrive at their decisions is crucial for gaining their trust and overcoming adoption barriers. Factory teams are siloed, autonomous, and operate under local leadership, making AI adoption challenging. Initial investment in AI systems and infrastructure can be substantial, depending on the approach, and many manufacturers may struggle to justify the expense.

AI relies on vast amounts of high-quality data, which may be fragmented, outdated, or inaccessible in many manufacturing environments. Legacy systems in manufacturing often run on vendor-dependent proprietary software, which use non-standard protocols and data formats, posing integration challenges with AI. Limited internet connectivity in remote locations requires overcoming latency challenges as manufacturing systems rely on accurate and reliable real-time response. For example, an AI system needs to process sensor data and camera images in real-time to identify defects as products move down the line. A slight delay in detection could lead to defective products passing through quality control. Additionally, manufacturing AI systems need to meet stringent regulatory requirements and industry standards, adding complexity to AIs development and deployment processes. The field of AI is still evolving, and there is a lack of standardization in tools, frameworks, and methodologies.

Transformative AI adoption requires commitment and alignment from both OT and IT senior leadership. OT leaders benefit by realizing that a connected, smart industrial operation simplifies work without compromising uptime, safety, security, and reliability. Likewise, IT leaders demonstrate business value through AI technologies when they understand the uniqueness of shop floor requirements. In fact, OT can be viewed as a business function enabled by IT. Integrating OT and IT perspectives is crucial for realizing AIs business value, such as revenue growth, new products, and improved productivity. Leadership must craft a clear vision linking AI to strategic goals and foster a collaborative culture to drive functional and cultural change.

While vision provides the why behind AI adoption, successful AI adoption requires vision to be translated into action. The AI CoE bridges the gap between vision and action.

Overview: The AI CoE is a multi-disciplinary team of passionate AI and manufacturing subject matter experts (SMEs) who drive responsible AI adoption. They foster human-centric AI, standardize best practices, provide expertise, upskill the workforce, and ensure governance. They develop a modernization roadmap focused on edge computing and modern data platforms. The AI CoE can start small with 2-4 members and scale as needed. For the AI COE to be successful, it requires executive sponsorship and the ability to act autonomously. Figure 1 outlines the core capabilities of the AI CoE.

Figure 1 AI CoE capabilities

The AI CoE should champion explainable AI in manufacturing, where safety and uptime are critical. For example, when an AI model predicts equipment malfunction, a binary AI output such as failure likely or failure unlikely wont earn trust with factory personnel. Instead, an output such as Failure likely due to a 15% increase in vibration detected in the bearing sensor, similar to historical bearing failure patterns would make people more likely to trust AIs advice. AWS provides multiple ways to enhance AI model explainability.

The AI CoE should partner with HR and leadership to upskill staff in the AI-powered workplace by developing career paths and training programs that leverage existing skills. GenAI solutions can help close the skills gap by showcasing how AI complements worker expertise. Leaders should emphasize how AI-enabled capabilities can free up time for complex problem-solving and interpreting AI insights. For example, Hitachi, Ericsson, and AWS demonstrated computer vision by leveraging a private 5G wireless network that could inspect 24 times more components simultaneously than manual inspections to detect defects.

The AI CoE ensures collaboration and joint decision rights between AI solution builders and factory domain experts. Together, they work backwards from business goals, breaking down silos and converging on AI solutions to achieve desired results. Additionally, the CoE acts as a hub to pinpoint impactful AI use cases, evaluating factors such as data availability, quick success potential, and business value. For example, in a textile factory, the AI CoE can leverage data analysis to optimize energy-intensive processes, delivering cost savings and sustainability benefits. Explore additional use cases with the AWS AI Use Case explorer.

Governance and data platforms are critical for scaling manufacturing AI. The CoE establishes policies, standards, and processes for responsible, secure, and ethical AI use, including data governance and model lifecycle management. AWS offers several tools to build and deploy AI solutions responsibly. The CoE develops a secure data platform to connect diverse sources, enable real-time analysis, scalable AI, and achieve regulatory compliance. This data foundation lays the groundwork for broader AI adoption, as demonstrated by Mercks manufacturing data and analytics platform on AWS, which tripled performance and reduced costs by 50%.

The AI CoE evaluates and standardizes AI and GenAI technologies, tools, and vendors based on manufacturing needs, requirements, and best practices. AWS offers a comprehensive set of AI and Gen AI services to build, deploy, and manage solutions that reinvent customer experiences. Scaling AI requires automation. An AI CoE designs automated data and deployment pipelines that reduce manual work and errors, accelerating time-to-market. Toyota exemplifies AI deployment at scale by using AWS services to process data from millions of vehicles, enabling real-time responses in emergencies.

The value of the AI CoE should be measured in business terms. This requires a holistic approach that is a mix of both hard and soft metrics. Metrics should include business outcomes such as ROI, improved customer experience, efficiency, and productivity gains from manufacturing operations. Internal surveys can gauge employee and stakeholder sentiment towards AI. These metrics help stakeholders understand the value of the AI CoE and investments.

Figure 2 Steps for building AI CoE foundations

Setting up an AI CoE requires a phased approach as illustrated in Figure 2. The first step is to secure executive support from both OT and IT leadership. The next step is to assemble a diverse team of experts consisting of shop floor personnel and AI IT experts. The team is trained in AI and defines the objectives of the CoE. They identify and deliver pilot use cases to demonstrate value. In parallel, they develop and enhance governance frameworks, provide training, foster collaboration, gather feedback, and iterate for continuous improvement. Integrating Gen AI can further enhance the CoEs content creation and problem-solving abilities, accelerating AI adoption across the enterprise. An AI CoE evolves over time. Initially, it can take on a hands-on role, building expertise, setting standards, and launching pilot projects. Over time, they transition to an advisory role, providing training, facilitating collaboration, and tracking success metrics. This empowers the workforce and ensures long-term AI adoption.

AI and GenAI technologies have the potential to create radical, new product designs, drive unprecedented levels of manufacturing productivity, and optimize supply chain applications. Adopting these technologies requires a holistic approach that addresses technical, organizational, and cultural challenges. The AI CoE acts as a catalyst by bridging the gap between business needs and responsible AI solutions. It fosters collaboration, training, and data solutions to optimize efficiency, cut costs, and spur innovation on the factory floor.

Artificial Intelligence and Machine Learning for Industrial

AWS Industrial Data Platform (IDP)

AWS Cloud Adoption Framework for Artificial Intelligence, Machine Learning, and Generative AI

The organization of the future: Enabled by gen AI, driven by people

Deloitte: 2024 manufacturing industry outlook

World Economic Forum: Mastering AI quality for successful adoption of AI in manufacturing

Harnessing the AI Revolution in Industrial Operations: A Guidebook

Managing Organizational Transformation for Successful OT/IT Convergence

The Future of Industrial AI in Manufacturing

Digital Manufacturing escaping pilot purgatory

Nurani Parasuraman is part of the Customer Solutions team in AWS. He is passionate about helping enterprises succeed and realize significant benefits from cloud adoption by driving basic migration to large-scale cloud transformation across people, processes, and technology. Prior to joining AWS, he held multiple senior leadership positions and led technology delivery and transformation in financial services, retail, telecommunications, media, and manufacturing. He has an MBA in Finance and a BS in Mechanical Engineering.

Saurabh Sharma is a Technical and Strategic Sr. Customer Solutions Manager (CSM) at AWS. He is part of the account team that supports enterprise customers in their cloud transformation journey. In this role, Saurabh works with customers to drive cloud strategy & adoption, provides thought leadership on how to move and modernize their workloads that can help them move fast to cloud, and drive a culture of innovation

Matthew leads the Customer Solutions organization for our North American Automotive & Manufacturing division. He and his team focus on helping customers transformation across people, process, and technology. Prior to joining AWS, Matthew led efforts for numerous organization to transform their operational processes using automation, and AI/ML technologies

Go here to read the rest:
Empowering Manufacturing Innovation: How AI & GenAI Centers of Excellence can drive Modernization | Amazon Web ... - AWS Blog