People, Tech, Business
Factory Technology Director
22+ years of experience in ERP system R&D for garment manufacturing production management.
As the industrial internet wave sweeps across the globe, traditional vertical industry management software that once flourished in specific historical periods now faces unprecedented survival challenges.
Take "YiGuan JingLing" (Clothing Management Spirit) and the "Agile Manufacturing MES System" that I was deeply involved in developing during my career as examples. The rise and fall of these products actually tell the story of traditional standalone/LAN software's painful evolution under the tsunami of cloud-native and SaaS transformation.
"The Vanishing Middle Link"—this proposition is not alarmist. When enterprises no longer need an isolated "management tool" but instead require a "digital foundation" that can connect to devices downward and support business collaboration upward, traditional software must either achieve self-redemption or be eliminated by the times.
Between 2000 and 2010, vertical industry software represented by "YiGuan JingLing" succeeded by solving the problem of going "from nothing to something" in informatization.
I still remember that winter in 2009. To develop the "Agile Manufacturing MES System," I led a 5-person team to live for a full year in a 2,000-employee garment enterprise.
We weren't writing code—we were understanding the industry. Every morning, I attended production scheduling meetings with the workshop director; during lunch breaks, I sat by the production line eating boxed lunches with workers, listening to them complain about which processes were most exhausting and which links constantly had problems; late at night, I sat in the office, looking at piles of production orders, thinking about how to make data flow.
This deep frontline experience made me profoundly understand: the essence of software is not code, but the digital translation of business processes.
The core pain point of that era was simple—inaccurate data, slow data.
The value of "YiGuan JingLing"-type products lay in: through C/S (Client/Server) and B/S (Browser/Server) hybrid architecture, centralizing data scattered across paper forms and Excel spreadsheets into a unified database. This seemingly simple "centralization" was revolutionary at the time.
In that era when Java talent was scarce and PowerBuilder prevailed, being able to implement business logic digitally was enough to build industry barriers.
The software products of that time were the "middle layer" of enterprise management, acting like an independent safe that sealed enterprise data in local servers. Clients needed this kind of "safe"—data security, access control, no production impact from network issues.
This "closedness" in that era was an advantage, not a disadvantage.
After entering 2010, traditional development models faced enormous challenges.
Traditional development frameworks (like SSH: Struts+Spring+Hibernate) appeared inadequate when facing increasingly diverse and complex user requirements.
Manufacturing business needs are infinite and ever-changing:
While traditional software development models are often "rigid." Each new feature requires the complete cycle of requirements analysis, design, coding, testing, and deployment. When R&D gets bogged down in unfamiliar simulated business scenarios, the evolution speed of production tools cannot keep up with business process changes.
Even though the "Agile Manufacturing MES System" won the Third Prize of Zhongshan City Science and Technology Progress Award, it later faced the dilemma of "weak promotion and poor market results."
Why? Because the times have changed.
Clients no longer need an isolated MES system. They need:
The traditional software "standalone/LAN" architecture appears inadequate in the face of these new requirements.
Early projects were mostly "siloed" constructions where data could not interconnect. When I was responsible for the "Guangdong Cooperative Medical Informatization Platform," I discovered that even with powerful core functions, if basic data remains in "siloed" construction, system value is greatly diminished.
Yet in the industrial internet era, data silos are unacceptable:
By 2025, 49 "cross-industry" industrial internet platforms already cover 16 provinces, driving manufacturing toward high-end, intelligent, and green development. This full-link connectivity capability is something traditional "siloed" software cannot match.
Current mainstream MES systems have formed a "cloud-native" technology trend:
In comparison, traditional software's "monolithic architecture" is like an old building that cannot be expanded—want to add a room? Sorry, you have to rebuild the entire building.
Facing iteration pressure, traditional software developers embarked on difficult self-redemption. This transformation was clearly visible in my polishing of the "Comprehensive Practice Platform" from V1.0 to V3.0.
| Stage | Core Logic | Pain Points | Redemption Means | | :--- | :--- | :--- | :--- | | Project Stage (V1.0) | Outsourcing model, passively responding to requirements | Endless requirement changes, always playing catch-up | Clarify boundaries between business requirements and software functions | | Pain Stage (V2.0) | Bogged down in simulated business development | Code accumulation, exploding maintenance costs | Attempt transition from traditional R&D to business-based platform | | Platform Stage (V3.0) | Production tool innovation | How to balance flexibility and efficiency? | Introduce business foundation platform, achieve 200% efficiency improvement |
The V3.0 platformization experience gave me a 200% efficiency improvement—but this was built upon infinite pain and lessons from V1.0 and V2.0.
The essence of this redemption is "de-tooling."
I gradually realized: true redemption is not in doing "management tool training"—making users learn the system's complex operations; but in creating a "teaching tool"—letting the system understand the user's business logic, and through flexible configuration and platform capabilities, rapidly responding to changes in business requirements.
The rise of industrial low-code development platforms provides technical support for this "de-tooling":
Through these platforms, enterprises can rapidly build production management, equipment monitoring, supply chain collaboration, and other applications, significantly lowering development barriers and costs while shortening application launch cycles.
This isn't cutting corners—this is a productivity revolution. When a platform enables business personnel who don't understand code to configure systems that meet their own needs, software truly achieves the leap from "tool" to "empowerment."
The "disappearance" of "YiGuan JingLing"-type products is not the disappearance of industry demand, but rather the evolution of carrier form. In the cloud-native era, the survival path for traditional software lies in:
Solving the "backward productivity" problem through platformization. V3.0's 200% efficiency improvement wasn't free—it came from lessons learned during V1.0 and V2.0 iterations.
Platform for platform's sake is a trap. Platforms must emerge from real needs, not architectural aesthetics.
Solving underlying technical bottlenecks for concurrent access and multi-source data integration:
By 2025, digitalization in garment manufacturing is bidding farewell to the "system installation" era. Hardware-level investments like equipment automation, line transformation, and workshop visualization, without data connectivity at the software level, still cannot solve the information silo problem.
Twenty-plus years of industry accumulation—deep understanding of education, manufacturing, healthcare—remains a core asset that AI and cloud-native cannot replace.
Why? Because understanding is more important than code.
Every great MES architect needs to spend time on the shop floor. Not visiting—living there. You need to see the morning shift change, watch the bottleneck form after lunch, feel the rhythm of the production line.
This "field sense" is something no AI model, no low-code platform can replace.
The "disappearance" of "YiGuan JingLing"-type products is essentially the death of traditional software's carrier form, not the death of industry demand.
In this era when the "middle layer" gradually disappears, only by achieving the leap from "software development" to "platform empowerment" can traditional products find new footing in the industrial internet landscape.
The "face" of digitalization is supported by technology, but the "lining" must be rooted in business.
Without process adaptation, a system is like a supercar without fuel—technologically impressive but going nowhere. Only when developers shift from "writing good code" to "understanding the industry," from "implementing functions" to "empowering business," can digital transformation truly take root.
Those invisible moments—when a bottleneck is identified and cleared before it causes a line stoppage; when material waste is reduced by 0.5% across thousands of units; when a scheduler can make an informed decision because the data on their screen actually reflects what's happening on the floor—these don't happen because you bought expensive software.
They happen because someone—somewhere—understood both the mathematics of data flow and the reality of manufacturing.
Agility is forged, not bought.
I'm Daniel Peng. 22-year industry veteran, still thinking about how to use logic and technology to make Chinese manufacturing smarter—one factory floor at a time.
If you're experiencing the pains of digital transformation or want to learn more about practical experiences in transitioning traditional software to cloud-native, welcome to connect. We're not just building software—we're finding a path to the future for Chinese manufacturing.
Based on one year living inside a 2,000-employee garment factory, discover why 73% of MES implementations fail and what truly separates 'electronic dashboards' from genuine agile manufacturing.
Read More