Product Design Studio
The "Digital Pumper"
& Industrial IoT Grid
Bringing the $50B Oil Patch Online—Without the Internet. An offline-first ecosystem that recovered 6% net profit margins for operators in cellular dead zones.
The "Unfair" Challenge
Our client, a fast-growing SaaS provider, serves the "long tail" of the oil industry—independent producers who manage thousands of low-volume wells but lack the IT budget of majors like Chevron.
The industry was running on 1980s technology. Pumpers drove hundreds of miles daily to manually gauge tanks, scribbling numbers on grease-stained paper tickets.
The real killer was the "Blind Spot." Between visits, leaks and theft went undetected. To make matters worse, 90% of these wells are in cellular dead zones, making standard cloud-IoT solutions useless.
// CONSTRAINTS
- ✕No Cellular Connectivity (Dead Zones)
- ✕Harsh Environmental Conditions (Permian Basin)
- ✕Legacy Hardware Compatibility
- ✕Zero-Touch Compliance Requirements
From Prototype to Acquisition
The "Offline" MVP
We proved we could digitize the pumper without changing their workflow. We built a React Native app that functioned as a complete local operating system, performing complex oilfield math on-device.
The Hardware Bridge
To remove human error, we introduced the "Silent Watchdog"—a custom LoRaWAN sensor network. We engineered a battery-powered device that wakes up, pings via long-range radio (penetrating steel tanks), and sleeps.
The "Shadow Twin"
We built a cloud ingestion layer that correlated human data (Run Tickets) with machine data (Sensors). The system now flags theft (sensor drops without tickets) and predicts leaks before they happen.
"Thin Edge, Fat Cloud"
We architected a hybrid system: A local-first mobile app for the humans, and a LoRaWAN mesh network for the machines.
The Edge (Hardware)
Custom LoRaWAN sensors wake up every 15 mins, ping fluid levels via ultrasonic pressure, and sleep.
The App (Offline-First)
A complete OS on the phone. SQLite local DB syncs via Merkle Trees when connectivity returns.
The Cloud (AI Twin)
Anomaly detection engine correlates sensor data with run tickets to flag theft and leaks.
> DETECT_LEAK: DELTA > 5.2 BBL/HR
> INTERRUPT_SLEEP_CYCLE
> BROADCAST_ALERT: EMERGENCY_PACKET_0xFF
> GATEWAY_HANDSHAKE: ACK
Solving the Sync Conflict
When a pumper is offline for 3 days, their local data drifts from the server.
We implemented a custom "Merkle Tree" sync engine using SQLite. The device acts as the "Source of Truth" for field data. Every edit is append-only—we never overwrite, we version. This allowed operators to "replay" a pumper's day to see exactly when data was changed.
- Conflict-free offline writes
- Full audit trail of every interaction
- Seamless background syncing
The Impact
"We stopped guessing what was happening in the field. This system turned the lights on."