Our Founder's Expertise Helped Build One of the First AI Systems for Infrastructure

In 1993, the U.S. Army Construction Engineering Research Laboratories was looking for a better way to diagnose equipment failures at military heating plants. Downtime was expensive. Manual troubleshooting was slow. They turned to artificial intelligence.

The system they built was called MAD (Monitor And Diagnose). It was designed to interpret sensor data, identify malfunctions, and recommend corrective actions automatically. The same basic concept now powers predictive maintenance systems across every major industry.

But there was a problem. Early AI systems like MAD did not learn from data the way modern machine learning does. They worked by encoding human expertise as logical rules. If a sensor reads X while condition Y is true, then the problem is Z.

That meant the system was only as good as the experts who built it.

Stan Briggs founded Control Associates in 1968 after six years at Bailey Controls, where he serviced boilers and control systems across Ohio's industrial plants. By 1993, he had 25 years of hands-on experience with instrumentation and controls.

The Army needed someone who understood how boilers actually fail in the field. Not textbook theory. Real-world knowledge: which sensors drift first, which switches malfunction most often, how to interpret readings that do not quite match what the manual says.

From the official Army technical report:

"The expert interviewed most extensively was Stanley Briggs of Control Associates, an independent consultancy specializing in instrumentation and control. Mr. Briggs personally has over 25 years of experience with boilers, and contributed information on all aspects of boiler operation and maintenance."

The report continues:

"Mr. Briggs gave very valuable information about different types of switches, most common causes of their malfunctioning, as well as the different manufacturers and their respective market shares."

His knowledge became the system's knowledge. The diagnostic rules that MAD used to identify failures, interpret control signals, and recommend corrective actions were built directly from his field experience.

MAD was an expert system, an early form of AI that predates the machine learning systems we see today. Instead of learning patterns from massive datasets, expert systems encode human knowledge as if-then rules that a computer can execute.

The Army's MAD system could monitor boiler operations, diagnose failures, identify inefficient operation, and recommend actions to optimize combustion efficiency. It ran on an IBM-compatible PC with 640K of memory.

These early expert systems were the precursors to modern AI in infrastructure. The same goal that MAD had in 1993, automated diagnosis of equipment failures, is now pursued with neural networks and machine learning. But the underlying principle remains: the technology is only as good as the domain knowledge behind it.

We are in the AI age. Predictive maintenance, automated monitoring, and AI-driven diagnostics are becoming standard across water treatment, manufacturing, and utilities. The technology has advanced dramatically since 1993.

But the fundamental insight from the MAD project still holds. AI systems need domain expertise to work effectively. Someone has to know which data matters, how to interpret edge cases, and what the numbers actually mean in real operating conditions.

In 1993, the Army understood this. They did not just build software. They sought out people with decades of field experience and made that knowledge the foundation of their system.

Control Associates has 57 years of that expertise. The same depth of knowledge that the Army needed in 1993 is what we bring to every project today.

The full technical report is publicly available through the Defense Technical Information Center:

• Title: Application of Expert Systems for Diagnosing Equipment Failures at Central Energy Plants
• Report Number: USACERL Technical Report FE-94/04
• Document ID: ADA276909
• Date: December 1993

The document is unclassified and approved for public release: View the full report

Stan founded Control Associates in 1968. By the time the Army came calling in 1993, he had 25 years of field experience. Today, three generations of the Briggs family continue the work he started.

The technology has changed. We now work with telemetry systems, environmental compliance monitoring, and modern control panels. But the foundation is the same: practical knowledge of how equipment actually behaves, built from decades of hands-on experience.

In 1993, that expertise helped build one of the earliest AI systems for infrastructure diagnostics. Today, it helps facilities across Ohio and Western Pennsylvania keep their systems running reliably.