All work

The demo below runs on sample data.

03 / 07Interactive demo

Remote inspections over sub-100 ms video with hazard detection

Remote inspection. Live remote inspections with sub-100 ms WebRTC video, AR annotations, Gemini and YOLOv8 hazard detection, digital checklists and one-click PDF reports.

Client
An industrial operator auditing distributed sites
Location
GCC
Year
2025
Services
Mobile & Field Apps, AI Agents & Workflow Automation
Industry
Quality, MES and inspection software for manufacturers
Stack
React, React Native, Node.js, WebRTC, Mediasoup, Socket.IO, Fabric.js, Gemini, YOLOv8, Docker
Status
Interactive demo
demo/remote-inspectionOpen full screen
Sample data. No client names or real records are shown.
1 / 1A working mockup with sample data. Click around, nothing is saved.

Overview

ShichifukuTekx built a remote inspection system that lets an expert at a desk guide a technician on site over live video. It uses WebRTC through a Mediasoup SFU for sub-100 ms latency at up to 1080p 60 fps, flags hazards with Gemini and YOLOv8, records checklist evidence and generates a PDF report at the end.

Where they started

  1. 01

    Every site audit needed a senior inspector to travel. Remote sites waited days for a visit, and urgent issues waited with them.

  2. 02

    Phone video calls were too low-quality to read a gauge, a weld or a nameplate, and they left no structured record behind.

  3. 03

    Inspection checklists were paper forms filled in after the fact. Photos lived on personal phones and were hard to match to the checklist item they proved.

  4. 04

    Writing the final report took longer than the inspection itself.

What we built

We designed two interfaces around two roles. Experts work from a desktop console. Technicians use a React Native mobile app built for gloves, glare and patchy coverage.

  • Live videoWebRTC through a Mediasoup SFU for sub-100 ms latency, with adaptive bitrate up to 1080p 60 fps
  • AI hazard detectionGemini and YOLOv8 analyse frames for leaks, corrosion and missing PPE, then overlay bounding boxes with severity scores
  • AR annotationsexperts draw arrows, shapes and text on the live feed with Fabric.js, synced to the technician through Socket.IO
  • Digital checklistspass, fail and review states, with mandatory photo evidence on selected items
  • Built-in communicationfull-duplex audio and chat for serial numbers and technical notes
  • Reportingone click compiles session data, AI findings and images into a PDF with a compliance score

Role-based access separates experts and technicians, and session status (online, busy) shows who is available to take the next call.

An analytics dashboard tracks open issues and compliance scores across sites, so recurring defects show up as a trend instead of a one-off.

The services run in Docker containers on Node.js. Read more about our mobile app development and AI agents work.

What changed

  • Sub-100 ms, video latency. WebRTC through a Mediasoup SFU
  • 1080p 60 fps, adaptive video. Quality adjusts to the network on site
  • 2 vision models, hazard detection. Gemini and YOLOv8 flag leaks, corrosion and missing PPE
  • One-click, pDF inspection report. Findings, evidence photos and compliance score in one document

Tell us what you are building

Book a 30-minute call or send us a project brief.

Emailsales [at] shichifukutekx.aePhone and WhatsApp+971 58 505 7791
OfficeOffice C1-1F-SF10837, Ajman Free Zone C1 Building, Ajman
HoursSunday to Thursday, 9:00 to 18:00 GST
Message us on WhatsApp