While reliable data connectivity and modern analytics technologies are essential for transforming raw data to insights, industrial organizations must create a culture of workforce empowerment by training users to leverage these tools effectively.

Learning Objectives
- Understand that AI-enabled data analytics training can help with cross-industry skills shortages.
- Learn about generative AI benefits and considerations for industrial use.
- Consider how structured workflows can improve enterprise-level monitoring and invest in upskilling for organization-wide industrial benefits.
AI-enabled data analytics insights
- AI-enabled data analytics training can help with cross-industry skills shortages.
- Generative AI benefits and considerations for industrial use.
- Structured workflows are needed to improve enterprise-level monitoring and invest in upskilling for organization-wide industrial benefits.
Data is a key enabler for achieving corporate objectives set forth by many industrial organizations, guiding productivity enhancement, optimization and decarbonization efforts. However, despite significant investments in data acquisition and storage for many companies, the ability to convert raw data to meaningful and actionable insights is far from reaching its full potential.
Modern analytics technologies provide a mechanism for transforming data to insights, but this requires a workforce capable of understanding the inputs and interpreting insights. As the industrial sector continues experiencing a significant drop in the average tenure and time-in-role for employees, it is rapidly losing experienced subject matter expert (SME) knowledge. Investments in new technologies like advanced analytics, AI, and machine learning are helping bridge this problematic knowledge gap.
Ensuring literacy in these emerging technologies requires organizational investment in training and upskilling the workforce, which can be difficult due to time and other resource constraints. However, technology partners are supporting and accelerating companies in these initiatives to help maximize user adoption and proficiency in these tools.
Cross-industry skills shortage: AI, data analytics training
According to the World Economic Forum’s Future of Jobs Report 2023, six in 10 workers will require training before 2027, but only half of workers have access to adequate training opportunities today. The same report found that in all industries collectively, training workers to use AI and big data ranks third among company upskilling priorities over the next five years.
Cross-industry surveys also are finding that large salaries are no longer enough to retain employees. Workers recognize that longevity in a sector and their future employability depend on access to and literacy in new technologies. For example, the 2024 Global Energy Talent Index Report found that 87% percent of workers in the oil and gas sector would consider switching jobs, with 50% citing professional growth and learning opportunities in a role as a key priority.
As supported by these reports, organizations that fail to adopt new technologies and enable their workforce to successfully use them are at risk of both losing their competitive edge and intensifying today’s glaring skilled labor shortage.
Advanced analytics and AI empower the industrial workforce
Fortunately, technology partners are simplifying procedures for organizations prioritizing modern technology adoption and workforce upskilling. As a result, these companies can expect measurable productivity and efficiency improvements, while also attracting and maintaining a talented and motivated workforce.
For instance, AI-equipped advanced analytics platforms enable teams to access multiple disparate data sources and seamlessly integrate, interrogate, and analyze the data regardless of its origin. This capability, in combination with intuitive self-service tools designed for time-series data analysis, empowers SMEs to transform their raw data into reliable, meaningful insights.
Access to these platforms enables users to make measurable impacts in their plants and, in some cases, across the broader organization, by completing root cause analyses, establishing baseline production models, and monitoring greenhouse gas emissions in real time. While the software is intuitive, users must be trained, both in the tools’ functions and features, and in the principles of data analytics, to maximize implementation benefit.
One of the greatest issues in technical training is a lack of time, especially in sustained blocks, to engage in formal training courses. Traditionally, training often required multiple days off-site, and its effectiveness was contingent upon numerous factors that were tough to control. To address these limitations, leading technology companies are pioneering new ways of upskilling the workforce, improving adoption with more time focused on value-adding activities.
Generative AI: Benefits and considerations for industrial use
One burgeoning example of companies’ investment in both technology and upskilling is the increasing adoption of generative AI (GenAI). This technology offers, among many other possibilities, a new opportunity for workers to obtain faster results, with industry-relevant, just-in-time learning, enabling teams to learn more efficiently. GenAI also can consolidate training and reference material, providing responses tailored to the specific equipment and processes of interest.
GenAI solutions provide a way to generate text or code based on a user’s prompt (Figure 1). For engineers without formal training or time to study programming languages like Python or R, accessing this feature within a Jupyter notebook environment significantly lowers the barrier to entry, and it facilitates better project understanding and collaboration with data scientist colleagues or third parties.
One national energy company found that by providing engineers with a GenAI assistant within their existing advanced analytics platform, they were able to accomplish a task in-house in just 15 minutes, which previously required more than four days and support from an outside coding team. Similarly, companies Ascend Performance Materials and British Sugar reported that AI assistance in the same platform has enabled their SMEs to cut analysis time in half, empowering teams to deliver more business value from their data.
When generative technologies are used to accelerate the implementation of well-known and understood techniques or methodologies, productivity gains are clear. However, when these tools are used to develop new or more complex use cases, providing access to these generative capabilities does not necessarily yield widespread adoption or satisfactory results.
Mistrust throughout industrial and commercial landscapes, stemming from early issues around hallucinations and inaccuracies, is difficult to combat. Without intentional and sustained upskilling programs and visibility and understanding of successful use cases, teams may be skeptical about dedicating their time to learning a new technology or wary of the results.
Therefore, it is essential for training to keep pace with technological development, including detailed explanations of rapid developments in prompt engineering, which teach the AI how to “think” about questions. Additionally, the retrieval of relevant documentation is driving exponential improvements in accuracy, increasing users’ confidence in the technology and providing the ability to trace results by verifying the data and input they are given.
It is also important for upskilling programs to discourage users from treating AI or generative technologies as a silver bullet to solve all operational issues. This is especially true in the process industries, which have an intrinsic need for data cleansing and contextualization due to the presence of sensor noise, measurement errors and drift, and distinct operating modes. Users must be equipped with knowledge to prepare their data effectively, as shown in Figure 2. Technology’s output is only as good as the quality of the data, and as the saying goes, garbage in equals garbage out.
When it comes to interpretation, end-users must also be trained on fundamental principles, such as how to validate results. When more workers are empowered to deploy complex machine learning algorithms using simple prompts, it is critical to precede that with the knowledge, understanding, and ability to evaluate the models’ validity. Pairing every new technological deployment with a comprehensive and well-defined training path is of the utmost importance.
Workers need sufficient training in these fundamental areas to achieve success with advanced machine learning techniques and AI implementations. Fortunately, remote and self-paced e-learning training, materials are more accessible than ever in employees’ daily workflows (Figure 3).
Offering these training formats empowers employees to learn analytics in their own style and at the time that is most convenient to them.
Structure workflows to improve enterprise-level monitoring
While generative functionalities, advanced analytics techniques, and scalable platforms significantly expand manufacturers’ production and productivity horizons, they also inevitably result in the need for new workflows. For example, this was the finding of Flint Hill Resources’ (FHR) monitoring department, who were tasked with remote surveillance of two refineries to identify early risks and proactively review asset health across thousands of unique assets.
To monitor the many thousands of alerts generated, the company needed a structured workflow so users could effectively triage, manage and monitor at scale. In partnership with the analytics software provider, FHR deployed an enterprise-level asset monitoring solution that enabled users to quickly evaluate whether action is needed based on any anomaly detected. The solution crucially provided the ability to capture feedback as well, ensuring models remain accurate and relevant (Figure 4).
For example, new operating conditions can increase the incidence of false positive alerts over time. Onsite engineers, with additional context and intimate process knowledge, can flag these alerts accordingly, then work with the monitoring team to reduce further occurrences. This collaborative workflow was essential in FHR’s monitoring objectives, empowering the company to efficiently monitor over 8,000 parameters and 4,000 assets.
When implementing analytics use cases at scale, users must have structured workflows to continually evaluate and provide feedback on the accuracy and reliability of the data insights they are provided, regardless of technique complexity: Be it a simple univariate threshold alert, or a complex machine learning algorithm. In addition to structure, this also ensures minimal mistrust, which can quickly derail technology implementation projects.
Invest in upskilling for organization-wide industrial benefits
While new technologies, such as advanced data analytics platforms and generative AI, offer countless opportunities to improve productivity, they also introduce new training and adoption requirements. Data remains the key enabler for optimization insight, but organizations must ensure their workforces are equipped with the knowledge to properly leverage, contextualize and analyze it.
Companies that commit to workforce empowerment and modern technological tools will experience the highest return on their investments, and opportunities to work with cutting-edge technologies support talent acquisition and retainment. This helps industrial organizations stave off challenges of the skilled labor shortage, driving efficiency, sustainability and profitability in continuously-evolving markets.
Fiona Guinee is a senior analytics engineer at Seeq. Edited by Mark T. Hoske, editor-in-chief, Control Engineering, WTWH Media, [email protected].
CONSIDER THIS
Are you using AI-enabled analytics for training, workflows and monitoring?
ONLINE
Control Engineering has other articles about AI and machine learning.