I'm doing engineering in the instance.

Chapter 1: The Wasteland

The moment his consciousness returned to the ground, Xie Chengzhou closed his eyes.

It wasn't out of fear, but because the correct procedure for entering an unfamiliar, enclosed space is: listen first, feel first, then open your eyes. Sight is the last reliable sense, especially when the lighting is uncertain. This is a habit he developed in underground engineering—in the rainy season in Bangladesh, in temporary structures in Malaysia, in that mine in Africa—pause for two seconds, firmly plant your feet, assess the ground's load-bearing capacity and the direction of vibrations, and then look.

He heard silence.

It wasn't true silence, but a silence with a background noise, the kind of silence where a building is breathing very slowly and very faintly. He put his feet firmly on the ground and felt the direction: low-frequency vibrations came up from the soles of his feet, from the northwest side, with a stable frequency and a cycle of about four to five seconds. It wasn't random; it was the rhythm of something continuously operating.

He memorized this: there is activity in the northwest direction.

Then he opened his eyes.

Abandoned chemical plant.

The three-story concrete frame structure appears to have been built in the 1970s, based on the relatively small column cross-sections. Design codes from the 1970s were more conservative in load calculations, using standards from that era. The cross-sectional width was approximately 350 millimeters, almost a full size thinner than modern industrial building frame columns. The roof is damaged, allowing natural light to stream in at an angle, creating several clearly defined lines of light and shadow on the ground. The light is westward, around 3 or 4 pm. There is a slightly pungent odor in the air, not from recent exposure, but from long-term accumulation of volatile residues in the concrete pores—this factory has been abandoned for at least twenty years.

Xie Chengzhou stood at the entrance without moving, first conducting a structural assessment.

His eyes scanned the area from left to right, from high to low.

A vertical crack, estimated to be over three millimeters wide, is present in the southeast corner column. Located in the lower middle part of the column, this is a typical failure pattern caused by load eccentricity, indicating a load shift towards the west-facing frame beams. The second-floor slab has two gaps: one with irregular edges, indicating natural aging and collapse; the other with neat edges, indicating external impact, not structural aging. The third-floor slab is not clearly visible from this angle, but the roof breach is on the north side, suggesting uneven load distribution above, with a heavier load on the north side.

Thirty seconds. He completed the scan in thirty seconds, mentally noting: the structure is relatively stable, the southeast corner pillar is the primary monitoring point, the north side is the risk zone, and it is temporarily usable.

Then he saw the orange-red area on the ground.

It was dry, crystalline, and irregularly shaped, with edges that appeared truncated—not the form of liquid evaporating naturally; the edges were too neat, as if it had spread along a structural fissure and then been cut off. He saved this detail without immediately starting the analysis, and continued scanning other areas.

On the west side of the factory walkway, in the northwest corner, there is a shadow moving.

He refocused his attention in that direction, pressed his body against the nearest pillar, and remained still.

About forty seconds later, he confirmed the outline of the object.

The figure was humanoid, about two meters tall, dressed in work clothes and wearing a safety helmet, its face obscured. It moved along the west side walkway, its steps heavy, each step spaced about two seconds apart, at a steady pace. He waited for it to complete one turn, glanced at the route, and then waited for it to complete the second turn—the route was fixed, not a random patrol, but a repetition of a defined trajectory.

Xie Chengzhou took the memo out of his waist bag.

Before entering this space, he had already checked the contents of his waist pack: a memo, a crack gauge, a pen, a watch, and a flashlight. These were items he always carried on real construction sites; the fact that he could bring them in indicated that this "realm" didn't have a mechanism to remove personal belongings, at least not light tools. He created a new page in his memo and wrote: "Site #001. Scene: Abandoned industrial building, three-story concrete frame, southeast entrance. Structural status: Generally stable, risk on the north side, southeast pillar is a key monitoring point. Threat entity: One, northwest side, patrolling a fixed route."

He then switched his watch to timer mode and began timing the threatening entity's lap time.

The thing took four minutes and fifty-three seconds to complete one loop. The route was an irregular rectangle covering the main area of ​​the factory's first floor, but it missed two blind spots—near the southeast corner pillar and an area of ​​about two square meters to the left of the north staircase. He sketched the route in his memo, marking the two blind spots.

Then he went to read the rules text.

The rules were already at the edge of his field of vision when he entered the space. He glanced at them but didn't read them carefully immediately. He wanted to figure out the routes of the threatening entities first before looking at the rules. This was his approach to projects: first look at the site, then the blueprints, and don't let the blueprints limit your judgment of the site.

Five explicit rules.

At the top of the rule text was a tiny header, about a third the size of the main text. He didn't recognize some of the characters; it looked like some kind of document numbering system. He tried to decipher it and only recognized three Chinese characters: "本构·荒场". He noted these three characters down in his memo, adding a note next to them: "本构: Meaning unknown, Source: Rule text header. To be verified." Then he continued.

He read it word by word, making mental notes.

"Rule 1: A factory supervisor patrols along a fixed route within the factory building. If detected, the experience immediately fails." — Confirmed, entity located, route recorded. "Rule 2: Experience termination condition: Find and activate the main switch in the third-floor control room." — Target confirmed, third floor, control room, main switch. "Rule 3: The factory supervisor's detection range is 10 meters. Beyond this range, they cannot be detected." — There's a problem here. He paused on the phrase "detection range." Detection range, not visual range, not auditory range, it's "perception range." The wording is vague. In engineering documents, vague wording either indicates carelessness or deliberate ambiguity. He wrote in his memo: "Rule 3: Detection method not clearly defined, pending verification." "Rule 4: The factory floor has limited load-bearing capacity; more than three people standing in the same area simultaneously will trigger a collapse." — Single person, not applicable for now, recorded for future reference. "Rule No. 5: The patrol route of the factory supervisor will be changed randomly every 30 minutes." — Variables, which need to be timed.

He reset the timer and started recording the start time of the current route from now on.

After reading the five rules, he summarized them as follows: The rules describe the limitations, but not the details of the boundaries. "A perception range of 10 meters" does not mean "everything within 10 meters will definitely be perceived," "the floor has limited load-bearing capacity" does not provide specific values, and "patrol routes change randomly" does not specify whether the change applies globally or locally. The rules leave a lot of ambiguity, which are either blind spots or traps.

His first step was not action, but verification.

He flipped to the new page in his memo and wrote a verification plan: "Items to be verified: ① Factory monitoring perception methods (visual? auditory? other?) ② Actual load-bearing capacity of the floor slab (What is the basis for the three people's assessment, structural or design rules?) ③ Scope and pattern of route changes (30-minute intervals, are there patterns in the direction of change?)"

He looked at the three items, circled "① Factory supervision perception method", and wrote a word next to it: "First".

What is a factory supervisor, and how does he perceive threats? These are the first variables he needs to confirm before taking action. Acting according to the literal meaning of the rules in the absence of sufficient data is the most common engineering accident logic he has seen—constructing according to the drawings without first verifying whether the drawings and the actual site are consistent.

He refocused his attention on the patrolling creature and began to think about how to validate the plan.

The orange-red area on the ground is right at the edge of the dead corner to the left of the north stairwell.

He had previously noticed the truncated edge feature of that area but hadn't analyzed it further. Now, after taking a closer look, he superimposed the "truncated" detail with the distribution of gaps in the floor slab structure and made a deduction: if the distribution of chemicals in this area spreads along the gaps in the floor slab structure, then its boundary is the boundary of the floor slab gaps. The existence of gaps means that the floor slab in this area is more likely to be hollow, and the sound transmission when stepped on will be different from that in normal areas.

He wrote this deduction in a memo: "Chemical area: edge cutoff, possibly related to floor slab gaps. Acoustic properties: to be tested."

Then he realized that he had just written the words "acoustic properties".

He paused there for a second.

"Acoustic properties" is a term he uses when inspecting underground spaces and enclosed structures. He wrote this term in this abandoned chemical plant not intentionally, but because the term automatically connected in his mind with the two inputs: "chemical area" and "floor gaps"—sound, conduction, range.

If the factory supervisor's perception method is not visual, but some kind of sound-related method, then the acoustic properties of the chemical area are not just a detail, but a key variable.

He added a line after "Acoustic Properties: To be tested": "Prioritize verification".

He closed the memo and placed his hand on the nearest pillar to feel the current vibration.

The low-frequency rhythm on the northwest side is still there, with a stable period and no change, indicating that the factory monitoring patrol is normal and no abnormalities have been found.

Xie Chengzhou stood still for about ten seconds to confirm the next action sequence, and then began to move towards the southeast corner.

Land on your toes first, slowly lowering your center of gravity, feeling the ground's feedback before each step, then shifting all your weight to it. He'd done this on soft foundations to determine if he'd sink; he'd also done it in African mines to determine the direction of ground vibrations. Now he was doing it on the concrete floor of an abandoned chemical plant, to determine if the sound of footsteps would fall within the detection range of something.

He took seven steps and then stopped.

There were no obvious abnormal noises from the ground, and sound transmission was normal, which was basically consistent with his expectation of the characteristics of walking on a concrete floor. He updated this information into his assessment: the route from the first-floor entrance to the southeast corner has low acoustic risk and is usable.

He noted down the time on his watch: about twenty-six minutes until the factory supervisor's last route change, and about three minutes and twenty seconds until it passed the southeast corner blind spot again. Three minutes and twenty seconds was enough time for him to complete the next step of verification preparations in the blind spot.

He continued moving towards the southeast corner, his steps slow and steady, listening with every step.

When he reached the vicinity of the southeast corner pillar, he first took another look at the vertical crack.

The crack extends downwards from the middle of the column, parallel to the column's axis. This is a typical characteristic of eccentric load failure, not shear or bending moment failure, but an early stage of pure compressive failure. The depth is estimated at about three millimeters, with the width greatest in the middle of the column, narrowing towards both ends—indicating that the failure started from the middle, and the load concentration point is near the location of the widest crack. He marked the area of ​​the widest crack in his memo: "Southeast column: Vertical crack, widest in the middle about 3mm, narrowing towards both ends. Load concentration point is in the middle; the load is shifting westward. There is no immediate acute risk, but prolonged leaning or application of lateral forces is not permitted."

Then he stepped back, leaned against the relatively intact pillar next to him, and glanced at the factory supervisor's current location.

It is currently in the western section of the corridor, with its back to the southeast corner. The route leads it along the north side of the factory building, about twenty-five meters from here. The sound is steady, the rhythm is unchanged, and it is in a normal state.

Xie Chengzhou flipped to the new page of the memo and organized the current list of information.

Known: Plant monitoring exists; route is fixed; sensing method to be verified; main target is the main switch in the third-floor control room; floor load-bearing capacity is limited; route changes every 30 minutes. To be verified: plant monitoring sensing method, acoustic properties of the chemical area, specific location and entry path of the third-floor control room. Available resources: southeast corner blind spot, north staircase (to be confirmed if this is the only available staircase).

He wrote a line at the end of the list: "Verification order: ① Plant monitoring and sensing methods (highest priority) → ② Acoustic properties of chemical areas → ③ Route planning → ④ Execution."

He then closed the memo and refocused his attention on the moving outline of the factory supervisor, waiting for it to complete its current segment before starting to time the interval between its next passage through the southeast corner.

He has plenty of time.

This wasn't the worst site he'd ever seen, not even in the top five. It was a three-story concrete frame, structurally stable, with fixed routes to the threatening entities, and while the rules were vague, there were only five. In Bangladesh, when he took over, there were seventeen variables to be addressed simultaneously, four of which contradicted each other, and a local subcontractor had quit on the third day.

On his memo, next to the line “Site #001 · Wasteland”, he wrote two words: “Controllable”.