She suddenly became aware that the drone next to her had its cameras trained on her, moving with her as she moved along the conveyor belt. Intrigued, she took a few steps forward, sideways, and back, to see if the cameras stayed on her. They did. Since she was the only moving object in the drone's visual range, she didn't think much about it. Not until the targeting module gave a faint beep and turned its status LED green. She recalled the information she received while visiting the plant with Dunwood.
"And when a drone locks on a target, this module lights green."
This drone had locked on her? She felt a chill down her spine, but tried to calm herself down. This must be a clear example of poor quality target acquisition performance. She approached the drone, looking around to see if anyone was approaching. She reached inside it and tried to remove its targeting module. She released the module brackets and gently pulled the module. It wouldn't budge. She pulled just a little bit harder and the module broke in little pieces, crumbling like a piece of cheese. The testing bank next to her turned a few LEDs solid red and beeped. She grabbed a fistful of module fragments and stuck them in her pocket. Then she moved away from there, hoping no one had seen her.
Outside in the parking lot, she took out her cell phone and called Tom.
"It's me. I need a technician, someone who can assess circuit boards and electronic modules in the field, and do it fast. I need him to look at a module as soon as possible. Uh-uh… OK, then, please let me know where he's coming from, and I'll meet him halfway. Great, I'll wait."
…78
Although checking the time every minute didn't help the technician get there any faster, Alex still couldn't help it. On her last day on this job, she was wasting a whole lot of time on a hunch, waiting for the tech to show up.
Preoccupied with her worries, she resumed her work, checking emails on the cloned laptop. Nothing much was happening, not by email, anyway.
"Excuse me?" The man seemed to have appeared from nowhere. "I am Josh Barnes from the lab. Mr. Isaac sent me; he said you needed my help?"
"Oh, yes, please sit down," Alex offered. "I need your help with this," she said, unfolding a napkin on the table and setting the pieces of broken circuit board on it. "I need any information you can give me, anything at all."
"How did it break?" Josh picked up a piece and studied it up close.
"I tried to pull it out of its mount. I released the brackets on both sides, and pulled gently. Not hard. I work in IT, so I know how to manipulate circuit boards."
"Show me how hard you pulled," Josh said, holding one of the bigger fragments with two fingers. "Pull at this piece, just like you pulled at the entire thing when you tried to remove it."
Alex grabbed the other end of the fragment and pulled. The fragment gave again, breaking into smaller pieces.
"Oh," Josh said, examining a fragment with a magnifying glass he took out of his toolkit. "That's interesting…"
"What?" Alex asked, feeling anxious.
"Well, from what I can tell, the entire circuit board is substandard. The thickness of the board itself should be significantly higher. For these complex modules, the manufacturer usually uses a reinforced board, so it can't become so brittle. This one almost crumbles in your hand. What's this supposed to do?"
"Did Tom have you sign an NDA? A nondisclosure agreement?"
"Yes, and he mentioned that you'd ask," he said.
"OK, then, these are circuit boards installed on UAVs."
"On drones? No way!"
"Yes, I'm afraid so."
"Something must be wrong with them," he said, scratching his head. "Electronic modules installed on equipment in motion is subjected to vibrations of all kinds. Therefore, they have exact standards of execution that are in place specifically to ensure that these boards don't crack, due to the shocks and vibrations they absorb."
"I think I follow," she said hesitantly.
"You see? Here," he said, pointing at the edge of the broken board fragment, "on the crack line, you can see there's not much coating applied. When a board is designed to work in a vibrating environment, it has to be coated with a special film, designed to increase elasticity of the board itself and to reduce the risk of coming apart or becoming loose from the original soldering. No such coating has been applied in this case. For the larger installed components on the boards, I was expecting to see vibration mounts, tiny little silicon or gel pads, to ensure that vibration doesn't harm the sensitive components. Takeoffs and landings pose a lot of mechanical stress on components, that's why circuit boards for aviation have special standards, designed to address this level of mechanical stress."
"I see," Alex said.
"You said you worked in IT? Well, I'm sure you can relate, then, if I can trigger your memory with this example. How are laptop components different from desktop components?"
"Ah, yes," Alex said, following his point.
"The laptop components are designed to work during motion, vibration, or even shock, as in dropping your laptop," he continued. "They have vibration mounts; the boards are coated to ensure they won't crack and parts become loose. Some components or subassemblies are even encased completely in gels or resins, to confer them the durability needed to withstand repeated shocks. This, however," he said, picking another fragment and studying it up close, "has none of that, or almost none. There are traces of anti-shock gel around this processor, right here, but nowhere else. This circuit board didn't spend enough time in the coating chamber. Are these boards even working properly?"
"Well, I'm not sure. Outside of breaking when trying to remove it, how else would this defect manifest itself?"
"Intermittent defects, that's what comes to mind. Before it breaks, it cracks. Let's say this board has a crack in it, from vibrations, a tiny little crack, barely visible, what we would call a hairline crack. When the cracked board is under stress, some of the circuit lines drawn on it would be interrupted, and the current would stop flowing through them. That would cause the board to malfunction, or completely shut down the system. After a while, when the board wasn't under stress anymore, these circuit lines would be restored, and the circuit board would resume proper operations."
"How would it malfunction?" Alex asked.
"It's impossible to predict. It varies from circuit board to circuit board and from crack to crack. It would also vary by the intensity of the stress applied and the duration of the disruption."
"So, for instance, could it cause a drone to be in and out of comlink with ground control, apparently with no reason?" Alex asked, thinking of the Florida incident. "Or, one minute it would obey ground control commands, then the next it would fly wherever?"
"Absolutely. It could have any type of intermittent defect you can think of. It could have entire modules shut down or function out of spec. The usual outcome for this type of defect is unpredictable, random, and on-and-off defects."
"Could it cause the drone to fire on targets on its own?" This time she was thinking of the Kandahar friendly fire incident.
"Absolutely not," Josh said. "Defects like these, driven by substandard components, can rarely cause a piece of equipment to start doing something it's not supposed to. Even if that happened, it would be minor, like turn on lights or something like that. Targeting vehicles or people and launching missiles on these targets, that's an entirely different story. It requires a complex set of instructions, which need to be driven by something or someone to make it happen." The tech obviously had put two and two together and was well aware of the Kandahar incident. Media coverage on the Kandahar friendly fire had been intensive; there was no surprise he was able to connect the dots on his own.