Chương 12: Chapter 11: Scavenger's Arithmetic

The production line completed its run at 0347 station time. One hundred and four connector pins, each one within Foundry Zero's specified tolerance of plus or minus fifteen microns. Ethan lifted the last batch from the collection tray, the tungsten-carbide parts clicking against each other like polished stones. The oil on his fingertips carried the sharp mineral scent of freshly cut metal, a smell he had not encountered in two years of archive work.

He set the connectors on the inventory shelf and turned to the terminal. Foundry Zero had been running calculations for six hours. The initial material audit — a complete cross-reference of Station Seven's inventory against the Precision Fabrication requirements — had expanded into something larger. The display showed a material dependency tree, each component branching into sub-components, each sub-component demanding specific feedstocks. Twenty-seven items in total.

The list glowed amber against the dark interface. CHROMIUM-MOLYBDENUM STEEL — 0 kg available / 47 kg required. TUNGSTEN CARBIDE BLANKS — 18 units / 120 required. SILICON WAFERS, 300MM — 0 / 24. HIGH-PURITY COPPER, 99. 99% — 3. 1 kg / 28 kg required. SYNTHETIC RUBBER COMPOUND — 0 L / 6 L required. OPTICAL QUARTZ BLANKS — 0 / 3. RARE EARTH MAGNETS, NEODYMIUM — 0 / 12. PRECISION BEARING STEEL, AISI 52100 — 0 kg / 34 kg required. LITHIUM ELECTROLYTE SOLUTION — 0 L / 14 L required.

VANADIUM ALLOY STOCK — 0 kg / 8 kg required. ELECTRONIC-GRADE SOLDER — 0. 2 kg / 4 kg required. The list ran on. Twenty-seven rows of requirements, twenty-seven columns of shortfall. Ethan pulled the station's colonial supply manifests from the archive. The records were two decades old — written before the evacuation, before the collapse, before everyone on this planet stopped answering their radios. Incomplete. Stations that had gone dark might have consumed their inventories in the interim.

Equipment might have degraded. But the manifests were the only map he had. Station Three: tungsten carbide in the industrial tooling inventory. Two hundred kilometers northeast. The station's status was unknown — no orbital handshake, no telemetry, no response on any frequency. A dead node. Station Nine: silicon wafers and optical quartz in the semiconductor cleanroom. Four hundred kilometers south. Also dead. Mining Outpost Epsilon: rare earth stockpiles, lithium compounds, vanadium-bearing ores.

Four hundred kilometers north. The most promising target on the list — a primary extraction site, not a manufacturing hub. Raw materials rather than processed components, but raw materials were what Foundry Zero was built to consume. Three locations. Fifteen missing items distributed across them, assuming the manifests were accurate and the equipment was intact and no one else had scavenged them in the twenty years since anyone last cared. Three assumptions.

Any one of them could turn the arithmetic from difficult to impossible. He worked the problem in reverse. Start with what he had — twelve of the twenty-seven items. The connector pins he had just manufactured. A partial supply of tungsten carbide. Copper wiring salvaged from decommissioned power conduits. Bearing steel might be substitutable — Foundry Zero could theoretically reprocess lower-grade steel through its smelting furnace, burning off impurities, adjusting the carbon content to specification.

But that required the smelting furnace to be operational. Which required the lithium electrolyte for its induction coil cooling system. Which was at Epsilon. The circular logic of scarcity. Every problem chained backward to the same missing link. Ethan spread the manifests across the workbench. Station Seven's inventory was the control group — he could verify what was actually here against what the manifest claimed was here. He had done that already during the initial survey.

The manifest was accurate within six percent. Acceptable margin of error for twenty-year-old paperwork. He applied that six percent confidence to the other stations. Station Three's manifest listed forty-seven kilograms of chromium-molybdenum steel in the industrial tooling bay. If the manifest was accurate, and if the station was intact, and if no one had touched it. Three conditionals. A probability tree with three branches, each one reducing the likelihood of success.

Probability analysis was an engineering discipline like any other. Define the variables. Quantify the uncertainties. Calculate the expected value. He defined three scenarios for each station. Optimistic: equipment intact, materials present, manifest accurate. Baseline: partial degradation, some materials recoverable. Pessimistic: station destroyed, materials lost, expedition wasted.

The optimistic scenario for Epsilon returned fourteen liters of lithium electrolyte, two kilograms of rare earth magnets, six kilograms of vanadium alloy, and assorted electronic components. The baseline scenario returned half that. The pessimistic scenario returned nothing and cost him whatever resources he expended to get there. Cost. That was the variable he had not yet quantified. He had no functional rover.

Station Seven's two surface vehicles had been parked in the vehicle bay since the evacuation — their power cells drained to zero, their tires cracked, their life support systems unmaintained. The diagnostic system reported one rover with a viable motor controller and intact suspension. The power cells were the bottleneck. Lithium-ion cells, military specification, designed to hold charge for decades. The cells themselves were functional — the internal chemistry had survived the idle period.

But they had drained below the minimum activation voltage. To recondition them, he needed to apply a low-current charge over several days, slowly bringing the voltage back to the point where the main charging system could take over. The charging system required lithium electrolyte for its thermal management loop. Lithium. Again. The rover could not reach Epsilon without lithium. The lithium was at Epsilon. Standard circular dependency — the kind of problem that required an input from outside the system.

He needed a transport mechanism that did not depend on lithium. The Silent Worker stood in Maintenance Bay A, its green status light blinking at quarter-second intervals. Awaiting production order. He pulled its specifications from Foundry Zero's database. The Worker was vacuum-rated to military standards — tested for continuous surface operation in Martian atmosphere, pressure differential tolerance of one hundred kPa minimum, temperature range from minus ninety to plus sixty Celsius.

Its power cell was a solid-state lithium-iron design, fully charged, with a continuous draw rating of sixty hours of mobility before recharge. Add portable solar arrays, and the operational window extended indefinitely — recharge during daylight, walk during daylight, hold position during night. If you encounter this story on Amazon, note that it's taken without permission from the author. Report it. The robot could walk to Epsilon. He ran the numbers. Four hundred kilometers to Epsilon.

Walking speed: three kilometers per hour on flat terrain, reduced to one point five on slopes or loose regolith. Average route: predominantly flat volcanic plain with one major elevation change at the Erebus escarpment. Estimated total transit time: one hundred sixty hours of walking. At ten hours of walking per day — constrained by solar recharge cycles — sixteen days one-way. Thirty-two days round-trip. A month of walking. For a robot that had already waited three thousand years for someone to give it an order.

The risk calculation was simpler than the rover analysis. The Worker had no life support to fail, no O2 to deplete, no human occupant to die if something went wrong. The primary failure modes were mechanical: actuator seizure from dust ingress, joint bearing failure, power system malfunction. The secondary failure modes were environmental: dust storm reducing solar efficiency below sustainable levels, terrain obstacle forcing route deviation beyond power reserves. He assigned probabilities.

Actuator failure: eight percent, based on the Worker's demonstrated reliability over five millennia of intermittent operation. Power system failure: four percent. Dust storm exceeding design parameters: twelve percent, adjusted upward from historical averages because the weather monitoring network had been offline for twenty years and prediction models were no longer reliable. Combined probability of mission failure: twenty-four percent. Roughly one in four. Acceptable.

The alternative — waiting at Station Seven for materials that would never arrive — had a one hundred percent failure probability. He spent the next two hours programming the mission parameters into the Worker's navigation system. The colonial surface topography database still loaded into Foundry Zero's memory provided elevation data at fifty-meter resolution. He plotted a route that followed the old colonial road network — the packed regolith of the transit corridors where dust was minimal and terrain was graded.

Waypoints at fifty-kilometer intervals where the Worker would stop, deploy solar arrays, and transmit a status report via the low-band emergency frequency. The cargo manifest loaded next. Primary objective: lithium electrolyte. Secondary: rare earth magnets, neodymium or samarium-cobalt. Tertiary: vanadium alloy stock. Quaternary: any electronic components with intact semiconductor substrates. The Worker's cargo capacity was forty kilograms — enough for the lithium and a few high-value items.

He wrote the mission logic. IF primary objective unavailable, search for secondary. IF secondary unavailable, search for tertiary. IF no objectives available, abort and return. IF encounter unsurvivable hazard, abort and return. IF power falls below twenty percent reserve, halt and deploy solar arrays. IF communication lost for more than seventy-two hours, continue mission autonomously and re-establish contact when possible. The last instruction was the hardest to program.

Not technically — the logic was straightforward. But it meant the Worker might vanish over the horizon and never come back, and he would sit in this station for weeks waiting for a signal that was never going to arrive. He programmed it anyway. The portable solar arrays required testing. He retrieved them from the emergency stores — three units, flexible photovoltaic panels on collapsible frames, military specification.

He deployed each one in the maintenance bay, connected them to the diagnostic system, and ran the full test sequence. Two units passed at ninety-four and ninety-six percent rated output. The third failed at forty-one percent — a micro-fracture in the panel surface, probably from thermal cycling during twenty years of storage. Two functional arrays. He mounted both on the Worker's back using the standard cargo brackets. The arrays folded flat during transit, deployed via a simple mechanical hinge. No motors.

No electronics. Just a latch mechanism that could survive anything Mars threw at it. He added an emergency repair kit to the cargo — spare actuator seals, joint lubricant, a small toolkit for field repairs. An emergency beacon, programmed to transmit the Worker's position on the colonial distress frequency if systems failed beyond recovery. If the Worker went down, the beacon would let him find it someday, when he had a rover. The morning of departure, he ran the pre-launch checklist three times.

Power cell: one hundred percent. Solar arrays: stowed, latches engaged. Navigation system: route loaded, waypoints confirmed. Cargo bay: empty, ready for loading at destination. Communication system: transmitting on schedule, receiving on schedule. Emergency protocols: loaded and confirmed. The Worker stood in the vehicle bay airlock, its green status light blinking at the standard interval. Ethan faced it across the pressure threshold.

"Mission summary," he said. The words came out rougher than he intended.

"Travel to Mining Outpost Epsilon. Primary objective is lithium electrolyte — fourteen liters minimum. Secondary objectives are rare earth magnets, vanadium alloy, electronic components. Load cargo to capacity. Return to Station Seven via the same route." The Worker's light blinked once. Acknowledged.

"Mission duration is approximately thirty-two days. Maintain communication schedule — status report every four hours on the low-band emergency frequency. If communication is lost, continue mission and re-establish when able." Another blink.

"If you encounter conditions beyond design tolerances — abort and return. The mission is not worth the platform." The Worker's speaker crackled. Its voice was flat, synthesized, the same voice that had said "Awaiting production order" to five thousand years of silence.

"Mission parameters accepted." The airlock inner door closed. The pumps cycled — atmospheric gases pulled back into the station's reserves, pressure equalizing with the surface. The outer door opened. Mars stretched to the horizon. Red regolith, pale brown sky, the sun a white disc overhead. The temperature reading on the airlock display showed minus fifty-three Celsius. The Silent Worker stepped through the door. Its feet pressed into the regolith, leaving prints in the fine red dust.

It walked north, its pace steady, its green status light visible against the red landscape. Ethan watched from the airlock control station. The Worker shrank with distance — first a human-sized figure, then a shape, then a dot. The green light was the last thing visible, a single pixel of color against the monochrome desert, and then it too vanished over the horizon. The surface map showed a green dot moving north at three kilometers per hour. All systems nominal. Mission underway. Estimated return: Day 32.

Eight days out, sixteen days searching, eight days back — he had revised the transit estimate upward after reviewing the Erebus escarpment topography in higher resolution. Thirty-two days. He turned from the display and walked back into the station. The maintenance bay was quiet. The Worker's charging station sat empty. The tools it had used to manufacture the connector pins were still laid out on the workbench. He picked up a connector. Turned it over in his fingers.

The surface was smooth, the tolerances within specification. A small thing. Twenty-six more items to go. The material hunt had begun. Mathematics dictated the pace now — probabilities and payloads, ranges and charge cycles, the cold arithmetic of restoring a civilization one kilogram at a time. He set the connector down and pulled up the station schematic.

The Worker was already fifty meters down the corridor, its green light receding into the darkness of the maintenance tunnel, its power cells fully charged, its task parameters loaded. In thirty-two days, it would return carrying lithium, rare earths, and vanadium — or it would not return, and Ethan would know that the restoration's first expedition had ended in silence somewhere on the volcanic plain. The station schematic glowed on the terminal. Thirty-two days of preparation.

He began with the micro-foundry's induction coil, which had been last calibrated in 2058.