[Life] 1997 Prototype of Intelligent Logistics

Wu Chao Hui (JEFFI CHAO HUI WU)

Article Date: June 25, 2025, Wednesday, 10:23 AM

In 1997, I left the printing company I had collaborated with for many years. That year marked the starting point of a complete change in the trajectory of my life.

I worked at the TNT customs clearance department in Sydney. At that time, there was no internet, no electronic dictionaries, and my English was not even professional enough; I could only rely on a thick paper dictionary, flipping through thousands of pages of customs codes. I arrived at the company at 5:30 AM to handle the customs clearance for express flights, and after work, I rushed to night school to study customs courses, getting home close to ten o'clock, taking a shower, and continuing to read late into the night. I almost translated the entire customs code.

The days of attending night school were one of the most densely packed periods of my life. For nearly a whole year, I left home early and returned late every day, working during the day, attending classes at night, and spending weekends in the library, with my sleep consistently falling short of four hours. It wasn't about a comfortable rhythm, but rather an extreme desire to understand the logic behind the system. Customs clearance wasn't about memorizing answers, but about understanding the structure; it wasn't about looking at experiences, but about mastering the pathways. And it was under this continuous pressure that I forcefully burned all the knowledge of customs clearance and coding rules into my brain, until all the customs clearance pathways formed an automated process in my mind.

It was an era when books had to be manually flipped and data had to be manually entered, with almost no modern "system support" available on the market. In the TNT customs clearance department, apart from the manager, there was only me. At that time, no one trusted a newcomer whose "English was not proficient enough," but I relied on learning, asking questions, and remembering—constantly consulting the manager, even if it meant getting scolded. Every time the manager taught me something, I would immediately record it on the computer and try to write simplified procedures myself, striving not to ask the same question twice. These early program segments greatly improved data entry efficiency and accuracy of judgment, and thus became the prototype for the "intelligent logistics system" I would later build.

This prototype is not a model based on theory, but one that is directly used in practical operations. It has been quietly running in the background, helping us handle hundreds of cases every day.

I still remember the most shocking moment, when the two of us completed nearly two hundred formal customs clearances in one day! This was not just about labeling, nor just filling out invoices, but the entire customs clearance process, including product coding, tariff calculation, additional tax assessment, and entry category review, all completed one by one. Even in today's logistics systems, it is extremely difficult to replicate this level of operational density under human limits.

My night school classmate from back then was astonished when I described this record, widening his eyes in disbelief and repeating, "Did I hear that right? Our company’s monthly volume is not even as much as what you do in a day!" In that era, completing twenty customs clearances was considered efficient, and companies that could handle fifty had basically adopted early automation templates. Meanwhile, my manager and I relied solely on manual labor and structure to achieve a task density far exceeding the industry average by dozens of times.

I later realized that it was not something that could be explained by "diligence," but rather that the structure itself was quietly completing an efficiency leap. For every operational process, I tried to break it down into independent modules and then combine them into a general template, gradually evolving an automatic reasoning logic path. I was not using high technology, but rather an extremely compressed process structure, elevating efficiency from "human operation" to "structure driving humans." This is the earliest intelligent logical framework.

In addition to the daily customs clearance process, I also proactively applied to rotate and learn in the sea freight customs clearance department, spending one to two days a week there without extra pay, just to master more case scenarios. The company surprisingly supported me, allowing me to advance in both express and sea freight, broadening my horizons and skills simultaneously.

Some customs clearance calculations are particularly complex, especially in DDP (Delivered Duty Paid) cases. The calculations involving multiple tax tiers, exchange rates, and conditions can take even an experienced manager 40 minutes to analyze, followed by another 15 minutes for verification; if any variable is incorrect, the entire process has to start over. However, my self-developed program only requires ten seconds to input basic parameters and can automatically output all duties, surcharges, and withholding/prepayment ratios. When the manager saw it for the first time, he was skeptical, but after several verifications, he was astonished and said, "This is even more accurate than our manual calculations." I knew that it wasn't inspiration, but rather the logical circuits I had built step by step over hundreds of late nights.

I worked at TNT for three and a half years without a single raise—because there were only two of us in the entire department, relying on this system to support all operations. In 1999, I followed my manager's advice to switch jobs, and my salary instantly increased by 50%. I later called to thank him, and he candidly replied, "Of course I hope for the best for you. After you left, the company struggled to fill your position with six people."

After that, I successively joined world-class courier companies such as DHL, UPS, and FedEx, where I encountered the actual operations of larger-scale and more complex systems, continuously refining the structural principles I had explored through practical experience. Joining the global Fortune 500 company Exel Logistics in 2000 was my first opportunity to validate my personal system design capabilities on a "large platform." I was no longer an executor, but a builder of system structures.

In Exel, I comprehensively integrated years of accumulated experience in customs clearance, calculation logic, cargo stratification, batch scheduling, and other processes to construct a truly meaningful "intelligent logistics system" original framework. It is not a product of software engineering, but rather an integrated logical architecture based on structural simplification + human brain simulation + spreadsheet modules. That year, I did not anticipate that this practice, starting from a paper dictionary + EXCEL, would become the genetic origin of the later "remote work system," "no-code scheduling platform," and "structure-driven supply chain management."

In that era without cloud storage, without automatic synchronization, and without AI models, I quietly wrote the prototype of the system into reality with a computer, a spreadsheet, and a brain.