Digital inspection platforms
Checklists bound to an asset, evidence attached, and a certificate at the end that a client can accept.
Detail →
Technologies
The stacks behind systems that run in the field
Dev Station builds digital inspection platforms and industrial IoT systems. These are the technologies those systems run on, taken from work in production.
40+Developers on staff
12+Years average leadership experience
6Stacks kept in daily use
3Markets served: UK, US, Australia
How we choose
A stack is a maintenance commitment
What we build has to keep working for years on a device somebody carries onto a site. That narrows the field fast. Where a client already runs a stack, we work in it, and only argue when the choice will cost them later.
Offline is the default
Every field app writes to the device first, then drains a queue the user can watch. Basements and lift shafts kill mobile data.
The instrument is part of the stack
Bluetooth LE, a gateway or the maker SDK. A reading off the tool beats a number typed from memory.
You own what we write
Repositories and cloud accounts stay in your name from the first commit.
01 / Stack
Back end technologies
Two back ends carry most of the work. ASP.NET Core sits under the platforms that started on the Microsoft side, Python with FastAPI under the newer services, and a broker behind both because readings arrive faster than a request cycle tolerates.
03 / Stack
Mobile technologies
Native iOS and Android where the app holds a Bluetooth session and a local database at once. Cross-platform through .NET MAUI, which is where Xamarin work now goes.
04 / Stack
Cloud technologies
Azure most often, because the platforms we took over were already there, with Azure SQL, Blob Storage and AD B2C behind sign-in. AWS and Google Cloud where the client already runs on them.
05 / Stack
DevOps technologies
Pipelines in Azure DevOps or GitHub Actions, containers where the load justifies them, infrastructure described in code. On a recovered project the first job is usually rebuilding the environments.
06 / Stack
AI technologies
Vision models for defect and presence checks, trained on your own images. Where a threshold and a rule would do the job, we say so and skip the model.
Decisions
Choices we have already made, and why
Ask any agency what they use and you get a logo wall. These are the positions behind ours.
1
Xamarin work now goes to .NET MAUI
Microsoft ended support on 1 May 2024. Two platforms we maintain started on Xamarin, so the migration is live work. If you run a Xamarin app today, the question is when, not whether.
2
Records are written to the device first
A system that needs a live connection at the moment of the check loses records exactly where the check matters. SQLite on the phone, a queue the user can see, and no closing a record without the evidence.
3
Instruments talk over BLE or their own SDK
Where the maker publishes an SDK we integrate against it, as we did with KTC for digital torque on the Hydrajaws platform. Otherwise Bluetooth LE or a gateway. Send model numbers and we will say which ones we can read.
4
Telemetry arrives over MQTT
Position and engine data arrive faster and less reliably than a request cycle tolerates. MQTT ingest with a broker in front keeps the console responsive when a hundred units report at once. Anything with a location goes in PostGIS.
What we build
What these stacks get used for
Industrial IoT systems
Sensors, gateways and the pipeline that turns a stream of readings into something a person can act on.
Detail →Field and driver apps
Offline-first apps for people working where the signal drops, with a queue they can watch drain.
Detail →Integrations into what you run
The ERP, the CMMS or the client portal that has to receive the result before the job is finished.
Detail →In production
Two systems and the stack each one runs on
Construction, United Kingdom
Work-at-height anchor and torque platform
React dashboard, ASP.NET Core on Azure, a .NET MAUI field app reading testers over Bluetooth, Power BI on top.
React, ASP.NET Core, Azure, .NET MAUI, BLE
Logistics
TruckFleet Labs
React 18 console, FastAPI behind it, PostGIS for stops and geofences, Redis for the queue, MQTT for telematics, native iOS for drivers.
React 18, FastAPI, PostGIS, Redis, MQTT, iOS
Our engineers have delivered for Karcher GmbH, Topicus Finance, Team Assurance and Standard Chartered Bank, alongside the equipment makers and inspection providers who make up most of the order book.
FAQs
What buyers ask about the stack
We already run on a stack you did not list. Can you work in it?
Usually yes. Most of our work starts inside a system somebody else built, and the first job is to document it before changing anything.
Our app is still on Xamarin. How bad is that?
Support ended on 1 May 2024, so you are running without updates until a store requirement or an OS release stops it. The upgrade path is .NET MAUI, and we have taken it on platforms we maintain.
Can you read our testers, gauges and torque tools?
Usually yes, over BLE or a gateway, or against the maker SDK where there is one. Send the model numbers and we will say which ones we can read.
Who owns the code and the cloud accounts?
You do, from the first commit. Repositories and cloud accounts stay in your name, so leaving us never needs a migration project.
Send us the stack you run now
Thirty minutes with the engineers on what you run, what it talks to and where it hurts. If the answer is to leave it alone, you will hear that.