The north side of Minneapolis once housed one of Minnesota's most concentrated collections of classified government documents. Most employees and visitors at Honeywell's Technology Center required security clearances issued to individuals on a need-to-know basis by various government agencies and registered by Honeywell's Security Department. I worked in that facility.

One of my first revelations about the critical confidentiality of secure documents came unexpectedly. One day, a month after I had presented a paper on the potential of an emerging technology at an International Sensors and Controls conference, the head of Honeywell Security visited my office unannounced. He informed me that he had a document addressed to me; however, I did not have the proper security clearance to read it.

He handed me a booklet of forms to complete and return to him as soon as possible. Not knowing what was involved, I said I would finish it by the end of the week.

Three weeks later, I finally returned the completed forms. To say they wanted to know every detail of my life up to that time makes it sound simple. But remembering and validating every home address, every application for a government license, every encounter with a foreigner, every relative who may have had a foreign connection, and every person who had paid me money during the previous 20 years took some concentration. I had to go back to the age of eight. I turned in the black-ink responses and then somewhat forgot about the effort.

I would later learn that several of my past and present neighbors had been visited and interviewed by intelligence officers. After four months I was visited by Honeywell's security chief once again, unannounced. He revealed that I now held a CIA top-level security clearance and could make an appointment with his office to open the envelope sent five months earlier.

Even today, after 40 years, I cannot disclose to you what was in that document. Suffice to say it led to significant new product developments for Honeywell that are still in use.

Hundreds of secret projects like this occupied our research center, each isolated on a need-to-know basis. Even our research director was not cleared for every project. The government agency involved issued clearances to individuals as necessary to complete the work. We signed affidavits acknowledging our commitment to handling documents and materials securely. Nearly every project developed prototypes of exotic, one-of-a-kind devices. Many could easily have been featured in James Bond movies.

Unfortunately, I can't tell you much about those devices. Some assisted captured service members in escaping. Others facilitated spying on an enemy in extraordinary ways. These projects were classified as Top Secret because disclosing them could have compromised agents in the field.

Some less experienced researchers sometimes questioned the need for so many cyber doors, locked cabinets, clean workstations and two-witnesses-at-all-times protocols for handling documents. Their questions would be coolly ignored. When it came to security with documents, there was no room for interpretation — only rules.

We were trained and retrained on security procedures. We never had a missing document, a document out of place, or classified data disclosed outside its security sphere. Not once in 20 years, involving hundreds of projects and thousands of individual clearances.

Well, there was one incident — which I didn't know about until a year after it had happened. I can give you some of these details because my part of the situation was unclassified, which became the problem's backbone.

I and one of my colleagues, Armin, devised a new way of using a sensor to detect submarines in deep water, a great concern to the U.S. Navy. We developed a model to predict the sensor's response under various scenarios. Ernie, one of our technology transfer managers with a previous nuclear submarine career, arranged a meeting with the Navy. My presentation seemed well received, although there were not many complex questions.

Two weeks later, I was visited for the third time, unannounced, by Honeywell's head of security. He entered my office with two investigators. They wanted every document associated with the research I had presented two weeks earlier — every overhead slide, every written report, every computer printout, every note, every missed-phone-call slip, every computer disk. They watched me erase every file on my computer related to the project and then took the laptop. Little was said. I was told not to ask questions.

The odd thing was that to me and Armin, whose materials they also seized, this work had seemed unclassified. We'd used public data to create our analysis. Public information on various submarines' dimensions and materials allowed us to analyze the distortion each vessel would make in Earth's magnetic field and then forecast the distance at which our sensor could detect that submarine.

After the seizures, I went upstairs to visit Ernie. The three security officers had also called on him. However, he had nothing to offer them as he had possessed none of our documentation. He was as bewildered as I was about the unannounced visit. I told him this seemed like a sign to stop working on the project. He agreed with a frown and a hint of a smile.

I would later learn that for the subsequent 12 months Ernie would be under investigation for violating Title 18 of the U.S. Code, Section 798. This law prohibits the willful transmittal of specified classified information to an unauthorized person. To be found guilty would mean the end of one's technical career, not to mention a possible 10-year prison sentence and $250,000 fine. Our presentation on the submarine detection idea had sent a shock wave through the U.S. naval security system.

The concern centered on the U.S. submarine data we'd worked with. Again, we had made our computations using data that was readily available in unclassified documents. However, many in the Navy felt no one could have accurately calculated the specific measurements we did, which they had believed existed only in Top Secret documents. We had not been aware of this mistaken belief of theirs, but Ernie had been. Or at least he was in a position to be aware of it, having been a captain of a nuclear submarine.

So the accusation was that Armin and I could not have come up with those calculations on our own. However, there was no evidence we had ever been exposed to the numbers in question. So there was reason to suspect a leak that had come from Ernie.

For 12 months, various U.S. intelligence agencies covertly circled our research center and any place Ernie had ever visited. If there was a "leak" it had to be identified and plugged with sudden and determined action.

Fortunately, for us, and especially for Ernie, the system was fair. In the end there was no evidence that Ernie had ever seen the numbers of concern, and thus no evidence he could have passed those numbers to us. No documents were found at his home or seen in his coat pocket. No one filed lawsuits to stop the investigation. No one saw classified documents lying around on Ernie's desk. He did not use his home as a storage facility for classified documents. If any of this had been the case, the response would have been swift and complete.

I disclose this story now to help others understand the seriousness we must employ in securing our classified data. We might think that some piece of classified data is not important. We may even consider that the information involved seems trivial, something we could discover or compute ourselves from open sources. But that's not how the system works.

It comes down to individuals and their commitment to keeping secret the classified information entrusted to them. Hundreds of Minnesotans have done it and continue to do so every day.

James Lenz, of Champaign, Ill., is a film producer ( and writes under the pseudonym T.H. Harbinger. He worked for Honeywell in Minneapolis from 1980 to 2002.