Chen Hao stared at the design drawing that was swaying incessantly in the wind, his fingers still gripping the vernier caliper. He didn't speak, but simply placed the caliper slowly on the table, making a soft sound.

Nana's screen was lit, and the 3D model of the new solution was still rotating. She waited a few seconds and noticed that no one was doing anything.

"Shall we begin assembly?" she asked.

Carl squatted on the ground, holding a section of the beam in his hand. He tested its weight, then put it down. "This thing can't hold up."

Susan stood on the other side of the workbench, writing a number on a scrap of material with a marker. She looked up at the window; the sky was overcast, and the wind was stronger than in the morning.

“Let’s try again,” Chen Hao finally spoke. “Let’s go with the new plan.”

The four of them gathered around the workbench again. This time, they didn't need to drill as many holes, and the structure was simplified, so theoretically it should be faster. But they only got halfway through the assembly when a problem arose.

When the polycarbonate sheet was placed into the slot, the edge got stuck. Carl pressed down hard, and the metal frame creaked.

"Don't force it," Susan said, reaching out to stop him.

Before he could finish speaking, the clips snapped open, the board slid off, and slammed onto the table with a dull thud.

"Here we go again," Chen Hao sighed.

They disassembled and reassembled it. The second time, they changed direction, first securing the pressure beam, then slowly inserting the mainboard. This time it went more smoothly, and just as they were about to close it, a gust of wind blew in through the broken window, making the entire frame tremble.

With a click, the right-side clip broke.

“The material is old.” Carl picked up the fragment. “This piece of metal is no longer usable.”

“It’s not that it’s bad, it’s that we didn’t choose the right one.” Susan laid out all the scrapped parts in a row. “Look, the problem is different every time. Sometimes the slot is too narrow, sometimes the beam is too light, and sometimes the board itself is warped. We’ve been patching it up, but we haven’t figured out what we’re trying to solve.”

The workshop fell silent.

Chen Hao sat down on the toolbox, picked up a bent metal strip, and twirled it around his finger again and again.

“I thought changing the design would solve the problem,” he said. “But it turns out we’re still stuck in the same place.”

“This is not your fault,” Nana said. “Under the current conditions, any single improvement is unlikely to overcome the systemic limitations.”

"What do you mean?"

“It means—” she pulled up the data panel, “that we are facing three problems happening at the same time: insufficient material performance, insufficient processing precision, and drastic environmental changes. If we only change one of these aspects, the other two will immediately become obstacles.”

Chen Hao nodded and threw the metal strip into the waste bin.

"What we're doing now is like using a colander to scoop water."

“The accuracy rate is 72 percent,” Nana said.

Susan chuckled: "You really think you can calculate that?"

“I’m serious,” Chen Hao stood up. “We can’t keep trying like this. Every time we try, we lose a piece of material. We need to figure out which step is the most crucial and get that done first.”

"How do we find it?" Carl asked.

"Look at the wall." Chen Hao pointed to the "car overpass wall."

The wall was covered with records of failures, densely packed together. Each case was followed by a reason and a conclusion. He walked over and looked at them one by one.

Test piece number one – the pressure strip was too thick, resulting in springback failure.

Number 2 – Insufficient frame rigidity, resulting in deformation under stress.

...

Number 7 – Stress concentration in the clip caused it to break instantly.

He looked around and then turned to Nana, asking, "Could you organize all of these? Find the places where the problems most frequently occur."

“It has been archived,” she said. “We are currently analyzing high-frequency failure nodes.”

Three sets of data popped up on the screen.

First, failures at connection points accounted for 68% of all cases;

Second, 57% of the sealed structures have insufficient weather resistance.

Third, low material compatibility leads to assembly difficulties, accounting for 49% of the cases.

“In other words, the clips and retaining strips are the biggest problem,” Susan said, pointing to the first one. “We always break down here.”

“And we’re using recycled materials,” Carl added. “The metal strips aren’t straight, and the plastic strips are brittle; forcing them together definitely wouldn’t hold together.”

“So it’s not that the design is bad.” Chen Hao stroked his chin. “It’s that we used scrap metal to match high-tech boards.”

"The root cause is a mismatch between resources and goals," Nana concluded.

"To put it another way, it means having lofty ambitions but a fragile life," Chen Hao grinned.

No one laughed.

He didn't care, turned around and walked to the bulletin board, picked up the chalk and drew a big circle in the blank space.

“In the past, we would try out any idea we had, and then revise it if we made a mistake. Now we’re taking a different approach.” He wrote three words in the circle: **Check, Analyze, Determine**.

"Investigate: List all the reasons for failure; categorize: see which can be changed and which cannot; decide: prioritize the most effective ones."

"When did you become so organized?" Susan asked.

"We were forced to." Chen Hao wiped the chalk dust off his hands. "If we keep messing around like this, we'll be blown apart by the wind before winter even arrives."

Nana began retrieving all the test logs. The electronic document refreshed page by page, while simultaneously being projected onto the wall. She marked the time, component, parameters, and operator for each failure.

Susan then began to inventory the remaining materials. She laid the metal strips out one by one on the ground, sorting them by straightness; she cut the plastic strips into small pieces, labeling them with their thickness and elasticity; she even separated the screws by length and degree of rust.

Carl wasn't idle either. He found some scrap paper and used a ruler to draw expansion simulation diagrams under different temperature differences. As he drew, he muttered, "Forty degrees during the day, below zero at night, this board expands and contracts twice a day, just like breathing."

Chen Hao watched the three of them work back and forth, and suddenly felt that something was different.

In the past, everyone was eager to get started, and whoever came up with a solution would go first. Now, no one is rushing to assemble things; instead, everyone is focused on organizing their belongings.

“Before, I wanted to win,” he said. “Now, I don’t want to lose.”

"Is there a difference?" Susan asked without looking up.

"Of course. Those who want to win always think about the final blow; those who don't want to lose will take every step carefully."

She stopped writing, glanced at him, and said, "This doesn't sound like you."

“I copied it from the wall.” He pointed to the bottom line of the “failure wall”—which he had just written yesterday: “Failure is not scary, what is scary is not knowing why you failed.”

Nana's data analysis is complete. Three red highlighted areas appear on the screen.

**Primary bottleneck: The connection structure cannot adapt to thermal expansion and contraction.**

**Secondary bottleneck: Lack of sustained sealing force in the sealing material**

**Potential Bottleneck: Inadequate Strength of Basic Materials**

“So,” Chen Hao stared at the screen, “we’ve been trying to make the clips hold the board firmly in place, but the board moves on its own. As soon as it moves, the clips snap open.”

“A flexible connection is needed,” Nana said.

"Is there anything readily available that can serve as a buffer layer?"

“There are some old tire scraps in the warehouse,” Carl said. “Cut them into strips and put them under the joints; they might absorb shock.”

"Let's give it a try." Chen Hao nodded. "Anything else?"

“The inner layer of the tent fabric has a waterproof membrane,” Susan said. “Tear it off and stick it to the seams; it will at least block the wind.”

"It sounds like patching up a hole."

"All we can do now is patch things up."

Chen Hao thought for a moment, then asked Nana, "Could we make a new structure? Without clips or screws, that automatically fits together in some way?"

“It’s theoretically feasible,” she said. “For example, using a double-layer structure with compressible material in between, and using the pressure difference between the inside and outside to achieve self-sealing.”

"That sounds even more mysterious."

“It’s like the lid of an old-fashioned thermos flask,” she explained. “Once you screw it on, a negative pressure is created inside, which naturally draws it in.”

“We don’t have a vacuum pump,” Karl shook his head.

“But we can create localized low pressure,” Nana continued. “For example, we can place tiny airbags at the seams and manually squeeze them to expel the air, creating an adhesive force.”

What are airbags made of?

“Rubber tubing and sealant,” she said. “If you can find a good bicycle inner tube, it will work even better.”

“There’s a bunch of old bicycles under the abandoned buildings in the West End,” Susan said. “We didn’t move them back before because the wheels were all flat.”

"Is it working now?" Carl asked.

"Maybe." Chen Hao's eyes lit up. "But let's not think that far ahead. The most important thing right now is to avoid using good materials to fill a hole."

He walked to the workbench and covered the remaining main materials with a dust cover.

“We’re not installing anything today,” he said. “For the next three days, we’ll only do two things: first, thoroughly classify all available materials; and second, based on the failure data, redefine the minimum standard—what kind of structure is considered acceptable.”

"For example?" Susan asked.

"For example, it can remain sealed for two hours in a level 10 wind; for example, the gap between the seams does not exceed 0.5 millimeters when the temperature difference between day and night is 40 degrees; for example, disassembly and assembly does not take more than 10 minutes."

"The requirements are quite high."

"It has to be high enough." Chen Hao slammed his hand on the table. "We're not making handicrafts, we're saving lives. If the wind blows in, people will freeze; if the rain leaks in, the electricity will break. There's no room for negotiation on this."

Nana has begun updating the database. She has entered the new testing standards into the system and generated a material compatibility table.

Susan posted the quick reference sheet on the bulletin board.

Carl drew a new assembly process sketch on the ground with chalk.

Chen Hao stood in the middle, watching all of this.

He knew that the frustration from earlier hadn't completely dissipated. Everyone was holding their breath, feeling that they shouldn't be in such a sorry state.

But he also understands that the real transformation does not begin with success, but with admitting defeat.

He was no longer in a hurry to prove anything.

He just wanted to figure out how to survive.

The wind is still blowing, and the blueprints are still swaying.

He went over and used a magnet to hold down the corner.

Then he turned to Nana and said:

"Put all the failure records back on. This time, we'll look at them one by one and disassemble them one by one."