By Danny Gavin, Communications Coordinator, Association of Equipment Manufacturers (AEM) --
As autonomous technologies gain traction across construction, agriculture, mining, and logistics, the non-road equipment manufacturing industry is moving to address an important challenge: how autonomous systems can operate safely alongside, and in partnership with, human decision-making.
Insights from large-scale deployments suggest that the “winning play” for autonomy does not eliminate the human role – instead, it elevates it.
“The organizations that succeed with autonomy will not be those that remove people from the loop but those that recognize human insight as essential to safe, reliable, and scalable autonomy,” said Mel Torrie, CEO of AEM member company Autonomous Solutions, Inc. (ASI), who presented on autonomy at the recent 2026 Product Safety & Stewardship Conference and Liability Seminar.
A Complex Safety Landscape
Despite rapid innovation, autonomous safety remains a work in progress. High-profile cases in the automotive sector have demonstrated that reliability and liability concerns are far from resolved.
These concerns have now reached the non-road equipment manufacturing industry as conversations around autonomy have intensified. Real-world deployments have shown that safety systems on equipment are frequently bypassed or overridden, particularly in operational environments where productivity pressures are high.
The takeaway? Safety is not just a technology challenge – it’s a human behavior challenge.
Learn more about AEM’s Safety & Product Leadership service area, and all the ways it helps AEM members and the industry maintain market access and comply with ever-evolving regulations
Lessons from Early Adoption
ASI’s early autonomy efforts across equipment types, from tractors to turf equipment, proved that the technology could work, but scaling it would be fairly difficult. Liability concerns, legal risk, and inconsistent real-world use cases limited broader adoption.
“When we did develop an autonomous tractor for one of our clients and went to launch it through their dealers, we had to have an awkward conversation with the company’s lawyers,” said Torrie. “Every year farms disable safety systems on purpose, and with only one hundred sales projected in the first year, we couldn’t cover the legal costs associated with the equipment.”
More success has been achieved in controlled environments, like in the mining industry. ASI has partnered with fellow AEM member company Epiroc to deploy a largescale autonomous fleet of mining equipment, which can be (and is) operated from hundreds of miles away, improving efficiency while allowing workers to avoid hazardous conditions and extended time on site. These systems succeeded through close collaboration with OEM partners, and an intensive focus on safety design and validation.
One key takeaway from these deployments is the importance of designing safety systems that can scale. Rather than developing one-off solutions for individual machines, leading organizations are shifting toward configurable systems that can be applied across entire fleets.
This shift enables greater consistency, simplifies deployment, and supports the standardization required as autonomy expands into more complex environments, including construction sites and logistics operations.
AI Is Accelerating Development and Raising the Stakes
Artificial intelligence (AI) is a driving factor that is dramatically accelerating autonomy development. Software can now be generated, tested, and deployed at unprecedented speed, lowering barriers to entry and intensifying competition.
As a result, competitive advantage is shifting. Software itself is becoming easier to replicate, while proprietary data and the insights it enables are emerging as the true differentiator.
Safety validation is also becoming heavily influenced by the technology.
“Advanced simulations can generate thousands of randomized scenarios across varying conditions, enabling far more comprehensive safety testing in far less time,” said Torrie.
One emerging practice of particular interest to the non-road equipment manufacturing industry is the use of agent-based safety testing. In this model, constructive AI agents validate expected system behavior, while adversarial AI agents actively search for gaps or unaddressed scenarios.
These simulations allow companies to test hypothetical stress conditions before they occur in real-world environments, strengthening system reliability and reducing risk at scale.
Keeping Humans at the Center
Even as autonomy evolves, human behavior remains a critical factor. In teleoperated systems, operators may be tempted to override safety alerts, particularly under time pressure. Designing systems that account for these behaviors by limiting overrides or adding additional safeguards is essential.
“We as an industry need to be looking at how easy it is to override safety features – maybe more steps are needed before users can skip these alerts,” said Torrie
At the same time, autonomy is reshaping other aspects of the workforce positively. Remote operations offer safer working conditions and improved quality of life, allowing workers to avoid hazardous environments and extended time away from home. Balancing these benefits with the broader workforce impacts of AI will remain an ongoing challenge for the industry.
The Industry’s Winning Play
Autonomy is not a short-term race. It requires long-term investment, continuous improvement, and a commitment to safety that extends beyond technology alone.
The path forward is becoming clear – autonomy’s success will depend on building scalable safety systems, leveraging AI responsibly, and designing each and every product with real-world human behavior in mind.
About AEM’s Product Safety & Stewardship Conference and Liability Seminar
The Product Safety & Stewardship Conference is setting the stage as the industry's only event designed to provide attendees with first-class access to the latest insights on product liability, safety design standards, regulatory requirements, and potential risks to avoid.
In addition to these core themes, the conference also explores emerging topics such as cybersecurity and AI, technical writing, electrification, and chemical compliance, ensuring a comprehensive view of the evolving landscape of product safety and stewardship.
Next year’s conference is set for April 19-22, 2027, in Bloomington, Minnesota. For more information, visit the official event website.
