Case Study · Industrial AR · Fortune 500

Semilynx

Fortune 500 factories were flying engineers across the country for every breakdown. We gave their technicians Google Glass — and eliminated it entirely.

60%
Downtime Reduction
38
Days to MVP
0
Cross-Country Flights Since Deploy
Stack: WebRTC Google Glass Azure IoT Python React
Semilynx Google Glass AR platform
The Challenge

$180K/month in avoidable engineer travel — and rising.

When a critical machine fails on a Fortune 500 factory floor, a specialist flies in. The round-trip alone costs $8,000–$15,000 before a single bolt is turned. For Semilynx's enterprise clients, this was happening multiple times per month across dozens of facilities.

Every hour of downtime cost significantly more than the travel itself. The real stakes: competitive manufacturing lines that couldn't afford to wait 24 hours for a specialist to land.

Why OnPoint Nexus

The only team that had shipped on Glass before.

Google Glass Enterprise development requires deep specialisation — most agencies had never touched the platform. OnPoint's AR engineering team had existing production experience and proposed a working prototype architecture on day one.

"Every other vendor quoted 6 months. OnPoint committed to MVP in 38 days — and delivered it in 36."

— Engineering director, Semilynx

What We Built

Remote expert eyes, live on the factory floor.

Live Video + Annotation

Sub-200ms WebRTC streaming from Google Glass to a remote expert console. Experts draw annotations directly onto the technician's field of view in real time.

Expert Remote Console

Web-based dashboard for remote specialists — live video feed, two-way audio, annotation tools, session recording, and machine diagnostic data overlay.

IoT Diagnostics Bridge

Azure IoT integration pulls live machine sensor data into the session — temperature, pressure, error codes — displayed as an AR overlay on the Glass screen.

The Hard Part

Achieving sub-200ms WebRTC latency from Google Glass — operating on patchy industrial WiFi — to a remote expert console. Factory environments are RF-hostile: metal structures, competing machinery, and variable signal strength. We built a custom adaptive bitrate algorithm and connection resilience layer that maintains session continuity even through 2-second signal drops.

The Results

Downtime cut by 60%. Zero flights since deployment.

  • ·60% reduction in factory downtime across all client facilities
  • ·Zero cross-country specialist flights since platform deployment
  • ·Average resolution time reduced from 26 hours to under 4 hours
  • ·99.97% platform uptime across 18 months of production operation
  • ·Expanded to 8 additional enterprise client facilities in year two
60%
Less Downtime
38
Days to MVP
99.97%
Platform Uptime
"
"Our engineers were flying across the country for every factory breakdown. OnPoint's Google Glass solution eliminated that entirely — and cut downtime by 60%. Remarkable execution."
DM

David M.

VP Operations · Semilynx, USA

Start a Project

Ready to build something like this?

If the problem is hard and the stakes are high, we're the right team. Discovery call is 30 minutes — no pitch deck, just engineers.