Customer Case Study · 150-Person Precision Manufacturing Company: One Network for Office, Workshop, and Warehouse

🏭 Client Profile

This is a metal products manufacturer in Shunyi, Beijing, with 150 employees and annual output value of over RMB 100 million. The workshop floor is densely packed with CNC machines, welding stations, and assembly lines, with a metal partition or steel support frame every few meters. With a tight production takt, workers rely heavily on the ERP system for barcode-based work reporting, material flow, and work order entry — the IT system has effectively become the “seventh process step” on the production line. The owner is not unfamiliar with IT; he simply does not want to be held hostage by it. Previously, annual spending on hardware procurement, outsourced operations, and network bandwidth was spread across three or four vendors, and the moment something went wrong they started pointing fingers — the network vendor blamed the slow server, the server vendor blamed the application developer’s poor code, and the application developer blamed insufficient bandwidth. No one took responsibility for the real cost of downtime. Mr. Wang, the company’s owner, later said, “What we fear most isn’t IT breaking down — it’s not being able to find anyone accountable when it does.”

🔥 Before We Partnered: Three Critical IT Problems

The first critical problem was the wireless network on the shop floor. Because the large amount of metal equipment strongly blocks Wi-Fi signals, barcode scanners deep in the workshop frequently lost connection at critical moments. Workers holding scanners and turning in circles searching for a signal was a common sight, and a single disconnection meant the entire workstation line paused for 5–10 minutes. During peak hours, such “minor glitches” could add up to nearly two hours of idle capacity in a single day. The second problem was the stability of the ERP system. Core business systems depended on LAN access, but the network architecture was chaotic — the office network, workshop network, and warehouse network were all mixed into a single broadcast domain, causing severe bandwidth contention. Whenever a large file transfer was underway in the warehouse, scanning latency on the production line spiked immediately. The production supervisor reported that in severe cases scanning failed to respond three to four times a day, each time requiring IT to step in and coordinate a restart. The third problem was the most serious: blurred lines of responsibility. One vendor handled hardware, another handled outsourced operations, and yet another handled network access. When a failure occurred, the company had no in-house IT professionals and simply could not determine which link in the chain was at fault. In 2014 there was a network outage that took a full 48 hours from report to recovery; during that time two production lines were forced to shut down for two days, with direct losses of more than RMB 100,000. Afterwards the three companies each issued their own reports and blamed one another, yet not one of them truly identified the root cause of the incident.

💡 What We Did

Step one: on-site survey and signal modeling (week 1). Our engineers spent two days on site, walking the workshop with a field strength meter and a laptop to measure signal strength distribution workstation by workstation. The diffraction effects of the metal partitions proved more complex than we had anticipated — the question wasn’t whether a signal existed, but which path brought the stronger signal. That site report became the basis for the design — AP placement was not decided on a hunch. Step two: re-planning industrial-grade APs (week 2). We added six industrial-grade ceiling-mounted APs in high-frequency scanning areas, using the 5GHz band to avoid 2.4GHz metal interference. Placement followed an overlapping coverage principle: even if one AP fails, the nearest AP takes over within 200ms, and the scanners never notice the switch. Signal strength across the workshop now holds steady above -55dBm, and packet loss dropped from 8% to under 0.2%. Step three: three-tier VLAN segmentation (weeks 3–4). We divided the network into three independent logical domains — office, production workshop, and warehouse. The office area runs on a gigabit uplink, dedicated workshop APs connect directly to the ERP server, and the warehouse RFID system runs on its own bandwidth channel. Access control policies between the three domains are enforced by Huawei S5720 series switches, completely isolating broadcast storms. Bandwidth contention disappeared at its root. Step four: unified operations and management delivery. We deployed a MikroTik CCR 1016 firewall as the egress gateway, configured intelligent link load balancing and QoS policies, and ensured ERP traffic always takes priority. At the same time, Siwenlide signed the service agreement as the single point of contact — any failure requires just one phone call. For the term of the contract, this equipment is provided to the customer free of charge; under the Network Equipment Package model, the customer no longer needs to purchase any additional hardware.

📊 Results After One Year

| Metric | Before | After | |—|—|—| | Failure response time | 48 hours | 30 minutes | | ERP scanning disconnections per month | About 30 | 0 | | Production halts per year | 6 | 0 | | Total annual IT spend | Spread across 3–4 vendors | One contract, about 38% saved | One detail behind these numbers deserves a mention of its own: in the year since this solution went live, cumulative downtime has not exceeded four hours, all of it during planned network cutover windows. Not a single production line has stopped for IT reasons since. Mr. Wang ran the numbers himself: the reduction in downtime losses alone covered the entire year’s IT service fees, with room to spare.

💬 What the Client Says

— Mr. Wang, Production Supervisor

When Mr. Wang said this, his pace of speech slowed noticeably compared with when he talks about other vendors. This company’s IT budget is not generous, and every yuan has to count. What truly made them feel they got their money’s worth wasn’t the performance of any single piece of equipment, but the fact that someone has their back when something goes wrong.

🤝 Best Fit for Similar Companies

If you meet any two of the following, this solution is very likely right for you:

1. 80–300 employees, with large amounts of metal equipment and partitions in your factory or campus that create complex wireless signal shielding;
2. Production processes that depend on barcode scanners, PDAs, RFID, and other mobile devices, where real-time responsiveness and accuracy directly affect production takt;
3. ERP, MES, or other core business systems that are sensitive to network latency, where an outage directly halts the production line;
4. IT vendors currently scattered across several providers — one for networking, one for operations, another for hardware procurement — who blame each other when problems arise;
5. Annual IT-related spending of more than RMB 150,000, yet no clear cost breakdown and no idea where the money goes.

Manufacturing builds IT on a completely different logic from internet companies. Internet companies chase the technological frontier; manufacturers want things that don’t fail when it matters most. What this solution solves isn’t making some metric look good — it’s turning the production system into infrastructure the owner can genuinely rely on.

📞 Get Started

Our engineers will call you back within 2 business hours. We start by understanding your current setup, then put together an assessment plan. No fees, no sales pressure: if it doesn’t feel like a fit once we’ve talked, just say so.