Manufacturing, energy, utilities, and aviation industries depend on experienced employees to maintain complex equipment, manage infrastructure, and keep daily operations running smoothly. However, as skilled workers approach retirement, organizations face a growing challenge: preserving the valuable knowledge and experience these employees have accumulated over the years.
The loss of institutional knowledge can affect productivity, operational continuity, and overall business performance. According to the U.S. Department of Energy, losing experienced employees creates significant risks for organizations. The challenge is particularly relevant in manufacturing, where the U.S. Census Bureau reported in 2020 that nearly one-fourth of workers were aged 55 or older. The aviation industry faces similar workforce pressures, with an estimated shortage of approximately 9,100 certificated mechanics in North America projected for 2026.
At the same time, maintenance problems and equipment downtime continue to create substantial financial losses. The National Institute of Standards and Technology estimated that preventable losses for U.S. manufacturers reached $119.1 billion, including $18.1 billion associated with downtime.
As these pressures grow, artificial intelligence is emerging as a valuable tool for helping industrial organizations preserve expertise, improve access to information, and make faster operational decisions.
In a recent episode of the Emerj AI in Business Podcast, host Eysan Yazici spoke with Vaibhav Kumar, Global Chief Technology Officer at IFS, about how businesses can use AI to transform enterprise systems and make operational knowledge more accessible. Their discussion highlights the importance of capturing employee expertise, simplifying access to enterprise information, prioritizing valuable data, and developing AI literacy throughout the organization.
Preserving Institutional Knowledge Before Experienced Workers Retire
One of the biggest challenges facing asset-intensive industries is the loss of experienced employees and the knowledge they take with them when they leave the workforce.
Long-serving technicians, engineers, and operators develop valuable expertise through years of working with equipment, solving problems, and managing unexpected situations. Much of this knowledge is not formally documented, making it difficult for newer employees to access the same practical insights.
According to Kumar, several pressures are affecting industrial organizations simultaneously. An aging workforce is creating skills shortages, operational inefficiencies continue to generate unnecessary costs, and supply chain disruptions are forcing companies to make decisions with incomplete information.
These challenges can have serious consequences. In industries such as aviation and critical manufacturing, a shortage of qualified technicians can delay maintenance and interrupt operations. Unplanned downtime can also result in substantial financial losses and regulatory penalties.
However, Kumar believes that the problem is not necessarily that experienced employees are unwilling to share their knowledge. Instead, traditional enterprise software often makes documenting and retrieving information unnecessarily difficult.
He explains that complicated software interfaces create friction for experienced workers who already have demanding responsibilities. When entering information becomes inconvenient, employees are less likely to document their expertise, even when that information could benefit the entire organization.
The same problem affects information retrieval. Employees may enter valuable operational data into enterprise systems, but if accessing and interpreting that information is equally complicated, much of its potential value remains unused.
Kumar believes that more accessible, consumer-friendly interfaces can help address both challenges. When employees can easily record their knowledge and retrieve useful information, organizations can create a continuous cycle of knowledge sharing. More contributions improve the information available to the workforce, while better answers encourage employees to continue using the system.
For industrial leaders, this means looking beyond technology adoption and examining how employees interact with existing systems. Identifying workflows where valuable knowledge is frequently left undocumented, along with records that are rarely accessed, can reveal opportunities where AI can make an immediate difference.
Giving Employees Direct Access to Operational Insights
Another important opportunity for AI in industrial environments is improving how employees access operational information.
Traditional enterprise systems often require workers to navigate multiple screens, generate reports, or request assistance from other departments just to find a specific answer. This process can slow down decision-making, particularly when information is distributed across several systems.
Kumar emphasizes the importance of creating role-specific experiences that employees can use without extensive training or unnecessary complexity. He believes conversational AI interfaces, similar to familiar consumer tools such as ChatGPT and Claude, can make enterprise information much easier to access.
By placing a conversational AI layer over existing business data, employees can ask questions in natural language instead of manually searching through complicated software.
Kumar illustrates this challenge through a common executive workflow. When a chief financial officer needs information from operational systems, the request may pass through several people before an answer is produced. The executive asks a vice president, who forwards the request to a manager. The manager then contacts an analyst, who retrieves the necessary information, prepares a spreadsheet, and creates a presentation.
By the time the information reaches the executive, it may not fully answer the original question, forcing the entire process to begin again.
With conversational AI connected to enterprise data, the executive could ask the question directly and receive an answer within minutes.
This approach makes operational insights more accessible to people who need them, reducing unnecessary communication layers and allowing employees to spend more time on meaningful work.
Kumar also shares his personal experience using AI to improve productivity. Tasks such as preparing presentations, which previously required several days of storyboarding, drafting, and review meetings, can now begin with an AI-generated draft in approximately 15 minutes. Additional refinement may take another 30 to 40 minutes.
The important distinction is that AI does not eliminate the person’s responsibility. The individual still defines the objective, provides the necessary context, reviews the results, and takes ownership of the final output. AI simply reduces the time between an idea and its execution.
Keeping Humans in Control of AI Agents
As enterprise AI becomes more advanced, organizations are also exploring agents capable of performing tasks with less direct supervision.
However, Kumar believes human involvement should remain central to how these systems operate. Employees should determine when agents can work independently and when they need to return for approval or further instructions.
This approach allows organizations to benefit from automation without giving up control over important decisions.
For industrial businesses, this distinction is especially important. AI agents may help retrieve maintenance records, coordinate workflows, or monitor operational information, but employees should still have the ability to determine which activities require human review.
The goal is to create systems that support workers rather than introduce another layer of complexity into their daily responsibilities.
Prioritizing AI Deployment Through Data Gravity
Successful industrial AI depends heavily on the quality and accessibility of enterprise data. Equipment histories, maintenance records, asset information, and service documentation provide the foundation that AI systems need to generate meaningful insights.
However, Kumar challenges the idea that organizations must achieve perfect data quality before introducing AI.
Many companies delay implementation because their information is incomplete, duplicated, or distributed across different systems. While data preparation remains important, waiting until every record is perfectly organized can prevent businesses from realizing the benefits of AI.
Kumar suggests that large language models and AI agents can also help identify duplicate records, recognize information gaps, and support data-cleaning activities.
He describes this approach through the concept of data gravity. Rather than attempting to apply AI everywhere at once, organizations should first identify the enterprise systems containing their most valuable operational information.
For example, a manufacturing company may have extensive data stored in equipment management systems, maintenance histories, production records, and service platforms. These systems contain information that can help employees understand equipment performance, identify recurring problems, and make more informed maintenance decisions.
By placing AI capabilities over these existing data sources, organizations can begin improving information accessibility without waiting for a complete data transformation project.
Kumar also encourages business leaders to evaluate AI products based on the actual capabilities they provide rather than their marketing labels. Some technology vendors may simply repackage existing integration tools and present them as advanced AI solutions.
For industrial organizations, the more important question is whether a solution provides a genuinely useful experience that allows employees to interact with business information naturally and efficiently.
The practical approach is to identify where the most meaningful data resides, explore opportunities to make it accessible through AI, and evaluate whether the technology delivers measurable improvements to existing workflows.
Building AI Literacy Across the Organization
Technology alone cannot guarantee successful AI adoption. Employees must also understand how artificial intelligence works, what it can accomplish, and where its limitations exist.
Kumar identifies AI literacy as one of the most important priorities for organizations preparing for the next stage of digital transformation.
He explains that public discussions about AI sometimes create unrealistic expectations or lead people to believe that the technology is not capable of delivering meaningful business results. However, many of the most significant advancements have occurred within the underlying technology infrastructure, while practical applications within enterprise software are still developing.
For AI to create measurable value, employees across the organization need to understand how to use it effectively.
This includes learning how context influences AI-generated responses, why starting a new conversation can matter, how AI skills are used, and how input and output tokens affect the way AI systems process information.
Importantly, AI education should not be limited to engineers, IT professionals, or employees directly involved in technology projects. Executives, managers, and frontline workers all need a basic understanding of the technology and its potential applications.
Kumar compares the current stage of AI adoption to the early development of cloud computing. Around 2010, cloud technology was still a relatively new concept for many business leaders. By the middle of the decade, executives had become more familiar with major cloud providers, infrastructure strategies, and regional redundancy.
He believes organizations are experiencing a similar transition with AI. Developing awareness and practical understanding today can help employees recognize where the technology can improve their work tomorrow.
Addressing Resistance to AI Adoption
Although business leaders are increasingly interested in AI, adoption does not always move smoothly throughout an organization.
Boards and executives may encourage companies to develop AI strategies, but middle managers and frontline employees may be more hesitant. Some worry about how automation could affect their positions, while others remain skeptical because previous technology initiatives did not deliver the expected results.
Kumar believes organizations should move away from the idea that AI is primarily a replacement for human workers. Instead, leaders should communicate how AI can help employees become more productive, reduce repetitive responsibilities, and focus on work that requires experience and judgment.
This shift in perspective is essential for building trust and encouraging participation.
Organizations that involve employees in identifying AI opportunities are more likely to discover practical applications that address real operational challenges. Executive support is important, but successful adoption also depends on making the technology useful and understandable to the people who work with it every day.
Kumar also warns that delaying AI adoption could create a competitive disadvantage. As AI becomes more deeply integrated into enterprise systems, it will increasingly influence employee expectations, productivity standards, and the way businesses operate.
Companies that begin investing in AI capabilities and workforce education today may be better positioned as these technologies become standard across industries.
The Future of Industrial Operations with AI
The conversation between Vaibhav Kumar and Eysan Yazici highlights an important shift in how industrial organizations should approach artificial intelligence.
AI is no longer simply about introducing new software or automating isolated tasks. Its greater potential lies in making existing enterprise systems more accessible, preserving valuable institutional knowledge, and helping employees turn operational data into meaningful decisions.
For industries facing workforce shortages, complicated software, and increasing operational pressures, these capabilities can make a significant difference.
Preserving the expertise of experienced employees helps prepare the next generation of workers. Conversational AI makes enterprise information easier to retrieve. Prioritizing systems with valuable operational data creates a practical starting point for implementation, while organization-wide AI literacy helps employees understand how to use the technology responsibly.
Ultimately, successful industrial AI adoption requires a balance between technology, people, and operational goals. Businesses must invest not only in intelligent systems but also in the data infrastructure, employee education, and governance needed to make those systems effective.
As AI continues to evolve, organizations that make operational knowledge easier to capture, access, and apply will have an important advantage in building more efficient, informed, and resilient industrial operations.