Why the Human Element Is Still a Necessary Aspect of Tunnel Monitoring

By Ellie Gabel

Brave professionals have dug and explored underground areas for as long as humans can remember. It is almost imperceptible to believe artificial intelligence (AI) will effectively work alongside humans in this sector, but this collaboration is becoming more normalized yearly. The joint effort will be critical for structural monitoring in tunnels and other subsurface structures, but it does not mean staff will become obsolete.

Contextual Understanding and Interpretation

AI and humans work together to understand the full scope of a tunneling project. However, each party excels in different areas. An AI is a swift pattern- and anomaly-identifier, parsing large data volumes in seconds. It could track temperature changes over history and how this alters soil quality in a specific area. Identifying specimens through computer vision and other advanced imaging is another advantage. 

It is a human’s job to contextualize and interpret the synthesized information. A miner, construction expert or engineer will transform the data visualizations into nuanced plans. For example, an AI could assess the best time of year to tackle a building or excavation project, but a project manager is the only one who can send the right staffers to the task based on their competencies. 

In an interview about these subjects, MIT doctoral student Michelle Vaccaro stated, “We found humans excel at subtasks involving contextual understanding and emotional intelligence, while AI systems excel at subtasks that are repetitive, high-volume, or data-driven.” The distillation summarizes where strengths lie and where they will excel in the foreseeable future.

Unforeseen Events and Adaptability

Predictive analytics is one of the most significant reasons corporations are adopting AI. While it is adept at discovering potential mishaps, AI is less reactive and intuitive than humans. 

Responding to a new situation would be impossible for a model, while a person could work through an equipment malfunction or geological unknown with more awareness. Humans can think critically through unforeseen circumstances, while an AI is only an expert at providing suggestions based on previous answers to a problem. 

If there is a cave-in or emergency scenario, a person may be able to find the best solution to help other staff, even though it would not abide by an AI algorithm’s programming. Workers can train an AI based on these adaptability learnings, but it takes significant time and effort. Alternatively, AIs have rapid hazard detection and 24/7 monitoring to support safety efforts without fatigue or delay.

Ethical Considerations and Accountability

Algorithms lack moral and ethical considerations, and internal accountability is only as strong as their training. An AI could generate recommendations on how to follow a legal requirement or face a safety concern, but it will not consider the emotional, philosophical and other quandaries unique to the human experience. 

Humans can use AI to brainstorm ideas, but they must translate how to make the ideas just and productive without introducing problematic behaviors. 

AI may notify workers of too many false alarms for structural breaches or climate threats. It only knows how to distribute these warnings based on programming instead of assessing the severity in context, which can take the workforce’s attention away from essential security problems or meaningful work. The technology could be an unintentional cause of heightened anxiety and alert fatigue, completely unaware of the emotional implications of its infrastructure.

System Validation and Continuous Improvement

People and AI make mistakes in different ways. In tandem, they can support each other in ongoing education and a growth mindset. Data scientists and engineers are essential for verifying an AI’s performance and maintaining informational integrity. 

If an AI reviews a tunnel, people must validate how well it represents the space. An AI inspecting an underwater highway tunnel repair project should be able to tell concrete professionals where blisters and microcracks are so they know how to prep the surfaces later. The positive feedback loop is critical.

Research shows robotics and AI are valuable for structural inspections when operated by humans through mixed reality headsets. The tests enhance defect detection without causing timing delays. The workforce leveraged the technology to boost their potential, and they were necessary for guiding it to success.

What Is Beneath the Surface of Human-AI Coworking

Many may fear AI will jeopardize job stability. When deployed based on its strength, AI is a powerful asset. It could increase safety, surveying potential, environmental awareness and navigational support, among many advantages. The human element will never disappear from tunneling projects, as its unique perceptiveness is crucial for careful exploration and construction.

Ellie Gabel is a freelance writer who specializes in covering the latest innovations in science and technology for an audience of industrial professionals. She is also the associate editor at Revolutionized.com.