Human-Centered AI and the Role of Design Thinking
Strong design work ensures that new technology supports human judgment rather than replacing it.
Artificial intelligence is rapidly finding its way into the utility sector, with industry organizations exploring how it can be used to identify hazards, analyze incident trends, summarize procedures and support planning activities. These capabilities offer exciting opportunities yet also prompt uncertainty. Today’s lineworkers, electricians, mechanics, operators and other trade employees are understandably questioning whether AI will eventually replace the experience and judgment that they have worked years to develop.
The short answer is no. AI will not replace skilled employees – assuming employers are thoughtful in why and how they deploy the technology. What does that mean? For starters, management must identify the employees most likely to use a prospective new AI tool long before rolling it out companywide.
Here is where design thinking becomes essential.
Systems vs. Design Thinking
Many readers are familiar with systems thinking, which helps us to better understand how work is performed in complex organizations. Design thinking is a human-centered approach to intentionally improving that work. When we investigate prospective AI tools, design thinking encourages us to assess how they can be used to boost task safety and ease.
This is not a new approach for modern safety professionals, who spend time each shift observing work, engaging with employees, identifying barriers and helping to refine processes. Design thinking merely provides a structured framework to do what many successful organizations already do well.
Consider the evolution of today’s vehicles, for example. Many now feature advanced systems engineered to automatically brake, maintain lane position, monitor blind spots and even assist with steering – technologies that have undoubtedly prevented crashes. Drivers can become increasingly reliant on these features, however, exhibiting slower reaction times if they malfunction. The problem here is not the technology itself but a flaw in how users and the technology were designed to interact.
Similar Challenge
Workplace safety currently faces a similar challenge. Imagine an AI tool that reviews job packages, weather conditions, equipment history, photographs and previous incident reports before work begins, with the goal of identifying hazards that could otherwise be overlooked and recommending risk controls. Used optimally, the tool is an invaluable second set of eyes; substandard use, though, could persuade workers to assume that no hazard exists where AI does not identify one.
Simply put, automation as a substitute for critical thinking increases risk exposure.
Design thinking aids in mitigating this risk by placing people at the center of the process, which ideally begins with management educating themselves about how experienced field crews recognize hazards, make decisions and adapt to changing conditions. Seek to understand crew members and their workflows rather than assume the details.
With insight into their employees’ everyday challenges, employers can develop potential solutions with AI’s assistance. Among its best applications is quickly executing otherwise time-consuming administrative tasks, such as organizing weather data, summarizing equipment history, identifying patterns based on previous incident data, and highlighting conditions that deserve further discussion.
Tool Testing Needed
Before adopting a prospective AI tool, recruit a small employee user group for testing purposes. Observe group members as they interact with it. Does the tool appear to encourage discussion and curiosity? Improve decision-making? Unintentionally reduce vigilance? Capturing and assessing this user feedback is essential to the decision-making process.
The best technology evolves alongside the workforce. This is an important distinction since work is rarely static; jobs can present varied and unexpected challenges in unusual and shifting environments. One of AI’s strengths is detecting patterns within large volumes of data, but experienced workers will often recognize subtle cues that data cannot fully capture. For instance, a seasoned lineworker will typically notice changing wind conditions before setting a pole. Veteran electricians sense when equipment doesn’t look right even though it has passed inspection. An astute mechanic tunes into abnormal vibrations before a monitoring system does. Keen professional judgment developed by industry workers through years of experience cannot be reduced to an algorithm.
Our real opportunity here is to amplify that expertise. AI can process thousands of past events in seconds; still, only experienced workers can determine whether the lessons learned then apply to real-time conditions now. Strong design thinking ensures that the technology an employer ultimately deploys supports human judgment instead of replacing it.
Three Guiding Principles
Three principles should guide organizational adoption of AI tools.
- Collaborate with frontline workers on implementation design. No software developer or AI model alone can fully replicate everything they know, so leverage this insight to shape your organization’s technology decisions from the start, well before any tool deployment.
- Attempt to solve meaningful problems the old-fashioned way before handing them over to AI. Digital assistants and other tech tools can diminish an individual’s need – and desire – to think critically about their job performance, increasing risk potential.
- Launch a small pilot program for each new AI tool that management considers adopting. Recruit, observe and gather candid feedback from a small employee test group. This data is especially useful when assessing a tool’s ability to evolve alongside its human users.
Conclusion
AI’s pervasiveness marks an extension of the utility safety professional’s role beyond hazard identification, incident investigations and regulatory compliance. We can uniquely position ourselves to bridge technology and frontline work by continuously seeking to understand how our company’s work gets done, how our employees recognize hazards, and where and when technology can strengthen situational awareness – or unintentionally weaken it.
The future of AI use in the utility sector will largely be shaped by how thoughtfully we integrate it with the realities of our work. Let’s strive to build or adopt tools that will enhance the knowledge and skills our workers already have, helping to ensure they get home safely after every shift.
About the Author: Gina Vanderlin, CSP, CHMM, CIT, CUSP, is the customer operations health and safety program manager at PSEG Long Island. With over 15 years of experience leading EHS initiatives in high-reliability industries, she remains passionate about elevating safety from a compliance function to a strategic driver of culture, engagement and operational excellence. Reach Vanderlin at gina.vanderlin@psegliny.com.