The Samsung Galaxy 7 fiasco may have begun with a tiny mistake on the assembly line

A version of this essay was originally published at Tech.pinions, a website dedicated to informed opinions, insight and perspective on the tech industry.

Those involved in the design and manufacture of hardware products understand that one of the most important phases of the process is testing. That’s the point when all of the assumptions that have been made need to be validated. The only way to do that is to build hundreds or thousands of units and subject them to a battery of tests. Even then, you might still find problems not anticipated once devices get into the hands of thousands of customers, but the goal is to be sure they are relatively minor.

The basic tests conducted include subjecting the products to a wide range of temperatures, humidity and physical abuse, including shock and vibration. The goal is to insure that the product performs the same before and after, and that the product remains intact and safe. Other tests include real-life user testing and measurements to insure that the product complies with regulatory requirements.

The testing typically takes several months to perform properly by a large group of quality and manufacturing engineers. Companies have rooms full of test equipment, including large ovens, shake tables and fixtures that exercise buttons and switches millions of times to simulate actual use.

Yet, in the case of the Samsung Galaxy Note 7, it is puzzling that the company claimed it was able to identify the problem with the initial shipment, fix it, test it and ship a half-million replacement units in just two weeks. That just doesn’t compute, and apparently, that suspicion was verified by the failures of the second batch of units.

So now, it’s quite possible the problem might have been caused by another component that interacts with the battery, rather than the battery itself.

Testing of smartphones is particularly important because batteries pack a huge amount of energy into a small volume. They contain circuitry to prevent a runaway condition should the battery or charging circuitry fail or go out of spec. The batteries are custom made to fit into the allotted space. Often, several companies or divisions are involved: The company building the battery cells, the company packaging the battery and adding the circuitry and connector, and the company putting the battery into the phone.

But here’s another opportunity for error: The company doing the assembly may have assumed that the battery integrator has performed sufficient testing. I’ve found that communications and clear division of responsibility among companies are often a weak point.

Yet in spite of a product passing all of this testing and having a sound design, there’s another thing to be worried about: How well the product is manufactured on the assembly line. Most lines rely on the use of many workers who perform the assembly operations, and not on automated assembly using robotic equipment. Each operator has instructions and tools to do a job that varies from attaching a circuit board assembly to the chassis, positioning and screwing the display in, or soldering a large component in place.

But it’s not uncommon for an operator to make a mistake: Not tightening a screw sufficiently or shorting out a circuit. To minimize this, other operators are interspersed in the assembly line to test the partial assemblies, and then the completed product will go through some functional tests to insure that it’s working.

But mistakes do happen. One electronic product I was involved in had a screw that was not tightened sufficiently. With little effort, it came loose and rattled around inside the product. That could be catastrophic, because the metal screw could short out a battery or blow a circuit. In this instance, the line was building 2,000 units a day on two eight-hour shifts, and by the time the problem was discovered, 8,000 units were affected. It was traced to one operator on one shift who failed to tighten the screw, even though she had a calibrated screwdriver that should have prevented this. So one individual who might have been distracted or wasn’t sufficiently trained caused a massive problem that required thousands of units to be opened, fixed and reassembled.

Imagine a factory building 100,000 units a day, and you can see how a small error can have huge consequences, much like the analogy of a butterfly flapping its wings and causing a hurricane halfway across the globe.

Phil Baker is a product development expert, author and journalist covering consumer technology. He has developed scores of products for companies, including Apple, Seiko, Polaroid, Barnes & Noble, Polycom, Proxima, ThinkOutside and Pono Music. Baker is the author of “From Concept to Consumer,” a former columnist for the San Diego Transcript and founder of Techsperts, Inc. Follow him at Baker on Tech, and reach him @pbaker.

This article originally appeared on Recode.net.