The skeleton in the vault appeared to have achieved partial success. Judging by his lack of hands and forearms, he’d gotten the lid open wide enough to reach inside – but he’d failed to prevent it slamming shut on him. A methodology that results in loss of limbs in this way has to be regarded as less than optimal. Leaving your fingerprints at the scene of a crime marks you as an amateur whose criminal career is likely to be short-lived. Leaving your actual hands behind will have even more drastic consequences. Bleeding to death being the most significant, I would say. I looked over at the skeleton and determined that I would take care not to suffer his fate.
Opening the lock on a safe, any safe, is only half the job. You have to prevent the safe from engaging additional bolts when it senses that someone is mounting an attack. Heat from a cutting torch or vibrations from an explosion cause a safe to activate hidden mechanisms that are independent of the main lock. If you don’t know what you’re doing, your joy at having cracked a safe’s lock will quickly turn to despair when you discover the door is still bolted shut. The purpose of these ‘re-lockers’ is to massively increase the time needed to open the safe once they’ve been triggered, in the hope that the thief will give up or take so long working on them that it gives the police time to turn up and catch him in the act.
The trick is to deal with these hidden bolts first, by either disabling the mechanism that triggers them or by blocking the bolts so they can’t lock into place. To do this you need to know how the mechanisms are constructed and where they are placed. There are a limited number of options and an experienced thief soon learns to recognise what he’s up against. As with most professions, we have to stay current, reading up on the latest technology. But my knowledge of the current state of the art was doing me no good on the Celestia. I needed to be an expert in vintage technology. Fortunately, modern mechanisms are often only more sophisticated versions of vintage ones.
When opening a safe, explosives are always a possibility. The noise involved is a disadvantage that has to be considered. But more significant is the risk of damaging the valuable items that you are seeking to extract from the safe. An artificial sentience is just about the most delicate bit of equipment you will ever come across. Even if you don’t blow it to smithereens, there is still a risk that the concussion caused by setting off a small bomb inside its storage box will damage it severely. I wouldn’t set off a bomb under my own hat, and similar consideration was needed here. I had to open the lock manually.
My first task was to get the top of the casing to pop open. For that, I needed to trigger the passive cooling system built into the box. And to achieve that, I need to generate heat. Much more heat than a blowtorch could manage. Something like the temperature of a small sun. Except, I realised, I didn’t. If Mr. Skellington’s twin brother had done this, there would be obvious scorch marks on the casing. But it had been pristine when I entered. Apart from the bloodstains. That meant there was an easier and less energetic way of triggering the lid.
The answer, like many such problems, was easy once I figured it out. Computers deal with data about situations rather than experiencing the situations themselves. If a machine receives data from a trusted source that tells it something is occurring, it accepts this as true. Unless programmed to do so, it won’t conduct independent analysis of the incoming data. If I could convince the computer that the temperature inside the box had risen to a critical level that threatened the continued existence of the Navigator inside, the ship’s computer would respond accordingly. That is, it would pop open the lid to provide physical ventilation.
“Trixie, is there any evidence of the Navigator communicating with the ship?”
“Still nothing.”
I wasn’t sure why the Navigator and the ship’s computer weren’t on speaking terms, but it made my task easier. If I told the ship that the Navigator was getting fried, she would ping a message to the Navigator asking for confirmation of the danger. If the Navigator said ‘No, I’m cool,’ my plan was doomed. If the Navigator failed to respond, the ship would trigger an emergency cooling protocol, fearing that non-response was due to the Navigator over-heating.
I tapped into the data link from the Navigator to the ship. I would send an ‘overheating’ signal to the ship, making it look like it had come from sensors in the Navigator’s box. I could then intercept the query from the ship to the Navigator and make sure it was never received or answered.
“Trixie, what’s the optimal temperature range for Navigator operation?”
“Minus forty to plus one-hundred-and-twenty degrees centigrade.”
I decided to increase the reported temperature slowly until it got to a hundred and then more rapidly after that. This would hopefully mimic a rapidly failing cooling system.
The result was disappointingly undramatic. There was a soft sigh and a gentle breath of air as the lid of the Navigator’s casing rose about a half inch. It was held up by a round silver bar at each corner, like a table leg. Stage 1 complete.
I used a small rotary tool with a blade attachment to cut some strips of metal
