Role of trusted data for AI innovation in manufacturing

By Mayank Baid, Regional Vice President, India, and South Asia, Cloudera

In the dynamic landscape of modern manufacturing, AI has emerged as a transformative force. It’s reshaping the industry, offering unprecedented efficiency and innovation. As India navigates the fourth and fifth industrial revolutions, AI technologies are sparking a shift in how products are designed, produced, and optimized.

Data is at the heart of this change. Armed with data from sensors, historical maintenance logs, and other contextual data, manufacturers can use AI to predict equipment behavior and potential failures, prescribe appropriate maintenance actions, optimise maintenance schedules, and reduce downtime. Additionally, AI can forecast product demand using historical data, trends, and external factors like weather and market conditions, generating immense value for manufacturers.

In India, the manufacturing sector is a significant contributor to the GDP, accounting for about 16-17%. With initiatives like “Make in India,” there is a strong push to enhance manufacturing capabilities, and AI is seen as a key enabler of this vision. However, whilst AI promises to drive smart intelligent factories, optimise production processes, enable predictive maintenance and pattern analysis, personalisation, and many other use cases, without a robust data management strategy, the road to effective AI is an uphill battle.

Understanding the value of data

Data—as the foundation of trusted AI—can lead the way to transforming business processes. However, many manufacturing executives say they are challenged when utilising innovations, including AI, for new use cases. According to a study by PwC India survey, 54% of companies in India are implementing artificial intelligence and analytics for business functions, however, 38% of the companies do not have plans to adopt digital technology for their businesses.

The lack of universal industrial data has been a major obstacle to AI adoption among manufacturers. While advanced technologies are a crucial part of the digital transformation story, manufacturers must first understand the role and value of data to excel in AI innovation. The low cost of sensors now allows manufacturers to collect, utilise, and manage massive amounts of data. However, if AI lacks access to a complete set of high-quality data, it will yield questionable analysis and suboptimal results. Organisations often build solutions on faulty assumptions, leading to biased, untrusted, and likely unsuccessful AI. Consequently, many organisations fail to realise AI’s value because they apply tools to inherently faulty data.

An automobile company has shifted its focus to the digital market to utilise its Big Data technology in new energy vehicles (NEVs) and Telematics. The rapid growth of NEVs has led to an exponential increase in vehicle monitoring data, challenging their data processing capabilities. To address these challenges, the company upgraded its Big Data Platform to a hybrid data platform. This upgrade significantly improved data storage and computation, reducing HBase storage by 73TB, cutting total cluster file count by 80 million, and boosting batch job performance up to 6.6 times. These improvements ensure detailed vehicle data analysis and compliance with national regulations.

Another such example of leveraging the data for the benefit of the company is of a tyre company that aimed to enhance fleet tyre monitoring to boost performance and safety. To achieve this, it migrated from an on-premises system to a hybrid data platform, benefiting from improved data management, machine learning, and analytics with high performance, scalability, and security.

Laying solid foundations

To combat data challenges and fuel data-driven AI in manufacturing, businesses must develop a data strategy built on a robust data platform. Collaboration between manufacturing operations and IT can foster a data-centric culture, enabling end-to-end data life cycle management focused on reliability and security, with a priority on data over complex AI systems.

Many manufacturing organisations in India still use legacy infrastructure and varied data sources on platforms like on-premise or public cloud. However, by deploying a holistic data platform built around a modern data architecture, manufacturers can eliminate data siloes by centralising data in a common data lake, offering the single source of truth that AI needs to flourish. This ensures AI is trained on or integrated with their own data, within their own networks and control, reducing the risk of data leakage and ensuring that AI outputs are contextual and accurate.

Realising the potential of AI

It’s clear that AI has the ability to revolutionise manufacturing. In India, this is especially crucial as the country aims to become a global manufacturing hub. But as with any new technology, there’s a risk of manufacturers focusing too intently on AI, without taking the necessary steps to ensure its success. Any AI implementation must be built on trusted data, underpinned by the solid foundations of a modern data architecture. Without this, organisations will fail to realise the true value of AI.

In an industry where even the slightest improvements can significantly enhance yields, those who harness the potential of AI will gain a substantial advantage, able to navigate the ever-changing manufacturing landscape. With India’s vast pool of skilled professionals and a burgeoning tech ecosystem, the country is well-positioned to lead the way in AI-driven manufacturing innovation.

AIITtechnology
Comments (0)
Add Comment