The Core UX Delivery: I stripped away the friction of diagnosing multimillion-dollar machinery over spotty, high-noise phone calls. I designed the cross-platform experience around factory realities—dirty hands, loud equipment, and poor reception. By combining freeze-frame markup (drawing directly on components), live sensor telemetry, and a store-and-forward data model that never lost state, I cut out verbal guesswork and gave operators and remote engineers a shared, indisputable diagnosis within minutes.
01. The Problem Space: The Cost of Industrial Stoppage
In high-speed commercial bakeries and industrial packaging facilities, uptime is measured in seconds. When a primary dough extruder, multi-tier industrial oven, or automated robotic packing line stalls, the facility does not simply lose production time; perishable materials spoil inside the line, creating thousands of dollars of losses per minute.
For decades, equipment manufacturers (OEMs) relied on an outdated service delivery model: emergency flights. When on-site maintenance crews could not identify a fault code, an OEM dispatched an expert field service engineer from Germany, Chicago, or Pennsylvania on a next-flight-out ticket.
The operational bottleneck was communication friction. Operators on the factory floor did not speak the specialized mechanical or electrical language of the equipment engineers. Commercial video conferencing software failed completely in this environment: cellular reception was shielded by heavy metal construction, overhead industrial machinery created ambient noise above 90 decibels, and standard video streaming protocols collapsed under volatile factory Wi-Fi networks.
02. Contextual Inquiry: Understanding Hostile Environments
Standard consumer product design heuristics fail on a factory floor. To understand how operators actually interacted with machinery during high-stress downtime emergencies, I conducted contextual research inside active manufacturing plants:
- Physical Affordances & Ergonomics: Operators wear heavy protective gloves and safety gear. Capacitive touchscreens respond poorly to greasy fingers, and multi-finger gestures or subtle swipe interactions are completely unusable while inspecting an oily drive assembly.
- Auditory Exclusion: Because ambient factory noise routinely exceeded 85–95 dB, two-way audio communication could not be the primary diagnostic channel. Spoken directions were regularly misunderstood or drowned out by nearby compressors.
- RF Shielding & Network Volatility: Metal cladding, electrical transformers, and heavy conveyors create severe dead-zones throughout modern production facilities. An operator stepping inside a machine enclosure routinely experienced immediate packet loss.
03. The Architecture: The Atomic "Packet" Data Model
To solve the challenge of asynchronous, low-bandwidth communication, I conceptualized and architected the atomic Packet data model.
Rather than treating diagnostic sessions as an uninterrupted, fragile video stream, Myi broke every troubleshooting event down into discrete, self-contained units of technical intelligence:
Synchronized Telemetry Bundles
A single Packet bundled together lightweight WebRTC video frames, compressed still images, high-contrast visual annotations, and instantaneous machine telemetry (spindle speed, motor temperature, drive pressure, and electrical fault registers). By pairing the visual state with machine sensor data, the remote OEM engineer received verified diagnostic ground-truth without relying on an operator's subjective explanation.
Deterministic Spatial Annotations
To bypass ambient noise, we developed freeze-frame spatial telestration. A remote engineer sitting at a desktop console in another state could freeze the operator's video feed, draw direct markup onto an electrical relay or pressure valve, and push that freeze-frame back to the handheld device instantly. The operator was not trying to hear audio instructions; they were looking directly at a highlighted component with directional arrows on their screen.
Store-and-Forward Resiliency
When an operator entered an RF-shielded enclosure with zero connectivity, Myi’s local caching engine continued capturing telemetry and high-resolution stills. The moment the device detected an accessible Wi-Fi or cellular beacon, it automatically negotiated a background reconciliation handshake, uploading queued Packets without data corruption.
04. Cross-Platform Execution: Balancing Web, iOS, & Android
As founder and design lead, I directed our engineering team of 4 through the complete architecture and delivery of three synchronized platforms:
- The Field Client (Native iOS & Android): Engineered for handheld durability, featuring oversized hit states, high-contrast daylight-readable UI states, one-tap voice-to-text memo toggles, and dedicated camera hardware button triggers.
- The Expert Console (Responsive Web): Built for remote OEM engineers managing multiple diagnostic sessions simultaneously, complete with split-screen multi-camera feeds, live machine register telemetry boards, and cross-session audit histories.
- Design-to-Engineering Prioritization: Managed sprint backlogs, authored technical API contracts between client and server layers, and established continuous integration checks to ensure WebRTC handshakes maintained latency under 180ms across unstable cellular links.
05. Commercial Impact & Industry Validation
Myi transformed commercial machinery support across Fortune 500 manufacturing facilities and global equipment OEMs:
- $6,000 Daily Cost Elimination: Saved enterprise clients an average of $6,000 per engineer visit day by resolving critical diagnostic failures remotely before dispatching field staff.
- Global Production Scale: Successfully deployed across 120+ enterprise facilities worldwide, supporting critical manufacturing lines for enterprises including Bimbo Bakeries and Sara Lee.
- 38,000+ Active Users: Maintained high adoption across plant floor operators and senior field engineers through intuitive, low-friction interaction design.
- Drastic MTTR Compression: Reduced Mean Time to Resolution (MTTR) on critical line-stopping electrical and mechanical errors from days down to under 45 minutes.