IoT & Connected Systems
Connected products fail at the seams — provisioning, connectivity, updates, and the moment data has to become a decision. We build the whole path: device firmware, edge gateway, ingestion, storage and the application layer people actually look at.
The shape of the work
Device to dashboard — and back again
Connected products fail at the seams. The return path — commands and over-the-air updates — is designed at the same time as the forward path, not bolted on after the first fleet ships.
Inside the IoT & Connected Systems practice
Device & edge software
Constrained-device firmware, edge gateways with local buffering, and store-and-forward behaviour for unreliable links.
Connectivity & provisioning
Cellular, LoRaWAN and Wi-Fi connectivity strategies with secure onboarding, certificate lifecycle and fleet identity.
Ingestion & platform
High-volume telemetry ingestion, time-series storage, rules engines and alerting built to scale with the fleet.
Fleet operations
OTA update campaigns, staged rollouts, remote diagnostics and device health reporting.
What we build with
Representative — we choose per project, and we will tell you when the boring option is the right one.
- MQTT
- LoRaWAN
- NB-IoT
- TimescaleDB
- InfluxDB
- Kafka
- AWS IoT
- Azure IoT Hub
- Grafana
Every engagement includes
- Security model documented from device to cloud
- Update path proven before first deployment
- Per-device cost modelled at fleet scale
What this actually involves
The level below the summary — for readers deciding whether we do this work or merely list it.
Connectivity, chosen for the site
The right radio is decided by the site, not by preference. A pumping station behind a hill, a farm across forty hectares and a factory floor are three different link budgets, and getting this wrong is expensive to undo.
- LoRaWAN for low-rate, long-range, battery-powered field sensing
- NB-IoT and LTE-M where cellular coverage exists and power allows
- Cellular routers with dual-SIM failover for gateways and RTUs
- Licence-free radio and mesh where there is no carrier coverage at all
- Wired fallback — Ethernet, RS-485 — wherever it is actually available
- Link budget and site survey before the hardware is specified
Edge autonomy — because the link will drop
The interesting design question is not what the device does when connected. It is what it does for the six hours it is not, and whether it lies about the gap afterwards.
- Control decisions taken locally, where the deadline is
- Store-and-forward with local buffering sized to the worst outage seen
- Timestamps applied at the source, so backfilled data lands in the right place
- Defined degraded-mode behaviour: what the device does alone, and for how long
- Safe state on power loss, brown-out and watchdog reset
Running a fleet, not a device
One device is a project. Ten thousand is an operations problem, and the difference shows up the first time you need to change something on all of them at once.
- Per-device identity issued at manufacture, provisioned without a shared secret
- Certificate rotation and revocation without a site visit
- Staged OTA campaigns — canary, phased, halt on failure rate
- Rollback to last known-good on a failed boot
- Remote diagnostics that characterise a fault before anyone drives out
- Battery, signal and health telemetry as first-class fleet metrics
From telemetry to something worth reading
Data volume is easy. The design work is retention, resolution and what a person is supposed to do when an alert fires at three in the morning.
- High-rate ingestion with per-device authentication
- Time-series storage with retention tiers and pre-computed rollups
- Rules and threshold engines with hysteresis, so alerts do not chatter
- Alert routing with escalation and acknowledgement
- Dashboards designed for an operator, not a demo
Other capabilities
Enterprise Applications
Line-of-business platforms, integrations and modernisation programmes for organisations that run on custom software.
Read moreData Centre & Infrastructure
Design, build and day-two operation of on-premise, colocated and hybrid-cloud infrastructure.
Read moreSCADA & Industrial Automation
PLC, SCADA and telemetry systems for plants, utilities and distributed field assets.
Read moreHave a project in IoT & Connected Systems?
Send the constraints. We will come back with an approach, a rough shape of the team and what it would take.