Flagship case study — Industrial fleet
4,073 silent machines.
Now they all speak.
A global manufacturer ran thousands of production machines on clipboards and guesswork, across 21 factories in 3 countries, with no visibility across sites. We wired every machine from scratch — custom electronics through live operations dashboard — and handed management a command center. Pilot live in 40 days.
instrumented
factory live
per day
3 countries
Running blind
at industrial scale.
When thousands of machines are invisible, you are not managing a production floor — you are guessing. This manufacturer had machines running around the clock across three countries and no way to know what any of them were doing until the next morning's report.
By the time the morning report arrived, yesterday's problems were already today's costs. A machine could sit idle for hours before anyone in management knew.
- Production data arrived 24 hours late — in spreadsheets, on paper
- No real-time visibility across factories in different countries
- Breakdowns discovered after production targets were already missed
- Manual headcounts to track output — slow, error-prone, one shift behind
- No way to compare efficiency across plants, shifts, or machine types
- Every factory running its own ad-hoc tracking patchwork
- Every machine visible, live — across all factories simultaneously
- Alerts within seconds of a fault, not hours or days later
- Management sees production, downtime, and efficiency in real time
- Automated output tracking — no manual headcounts, no morning-after lag
- Cross-plant, cross-shift benchmarking built in from day one
- One unified system replacing the entire patchwork operation
From factory floor
to command center.
X402 designed and built every layer of this stack — from the electronics bolted to the machine, to the dashboard on the factory manager's screen. No subcontractors. No handoffs. One team owns all five layers.
All five layers designed, built, and operated by X402 — custom electronics through production software
The fleet,
as it runs.
21 factories. Every machine reporting to the same platform, every hour. Each node below represents a live installation — amber means running, red flags an event the platform has already detected.
We don't outsource
the hard part.
Most integrators plug off-the-shelf sensors into a third-party platform and call it done. X402 started at the hardware layer and built everything up — because we had electrical engineers on payroll who could.
Custom electronics & firmware
Industrial machines don't have sensor ports. Our electrical engineers designed custom PCBs that tap into existing electrical signals — non-invasively, without interrupting production. Firmware handles local buffering, edge filtering, and compressed transmission. The hardware survives industrial environments: electromagnetic interference, voltage spikes, temperature extremes. When a machine at a remote plant goes dark, the alert reaches the same engineers who designed the circuit board.
⚡ Electrical engineers on payrollEncrypted mesh network
Connecting 4,000+ edge devices across factories in three countries — each on different network infrastructure — required a purpose-built mesh. Open-source encrypted tunnels, fault-tolerant routing, offline buffering at the edge. Data arrives at the platform even when a remote factory's internet link drops temporarily. No single point of failure between any machine and the central platform. All traffic end-to-end encrypted from the edge agent outward.
Platform & AI layer
The platform ingests 80,000+ signals per day and turns them into insight: real-time OEE (overall equipment effectiveness), downtime root-cause classification, production forecast vs. actuals, anomaly detection. We built a custom MES/ERP that replaced the client's entire patchwork of spreadsheets and legacy tools. AI models route alerts intelligently — a vibration spike that looks like a fault but isn't gets filtered; a genuine production bottleneck surfaces immediately to the manager on duty.
Operations & ongoing support
We run the production environment — cloud infrastructure, edge device OTA updates, platform uptime, data integrity. When a machine at a remote factory goes dark, the alert doesn't go to a support ticket queue — it goes to the same engineers who built the hardware and wrote the firmware. Mean time to resolution measured in minutes, not days. The system has grown continuously since the pilot: new factories onboarded, new machine types instrumented, new analytics modules shipped.
Numbers that moved
the business.
From contract to first factory live in 40 days. From one pilot plant to a 21-factory global fleet in under two years. The system now runs the entire operation.
The only team
that could do all of it.
This wasn't a job for a sensor vendor, a software house, or a systems integrator. It needed a team that could own the electronics, the firmware, the network, the platform, and the operations — and answer for every layer at 3 a.m.
- No data ports on existing machines — required custom hardware design from zero
- Industrial electrical environment: EMI, voltage spikes, extreme temperatures
- Connectivity across 3 countries with inconsistent infrastructure at each site
- 4,000+ edge devices that all need firmware updates, monitoring, and field replacement
- Platform had to absorb an entire MES/ERP — not just add sensors on top
- 40-day pilot target was non-negotiable — project had no second chance
- Electrical engineers on payroll — designed the hardware from scratch
- Embedded firmware team with production scars in industrial environments
- Network team that has run global infrastructure across multiple continents
- Platform team that built a custom MES/ERP from zero, owned the migration
- Operations team that runs the system in production — every day, indefinitely
- 40-day pilot delivered. 21-factory rollout completed.
Is your factory
flying blind?
If your production data is still living in spreadsheets — or arriving a full day late — we know exactly what needs to happen. Tell us about your operation. Honest answer within a day.