Solutions · IoT & Device Management
Connected hardware only pays off when the software behind it holds up. We build the provisioning, telemetry, dashboards, update pipelines and integrations that keep a fleet running for years, not months.
You own the code and IP Weekly demos Reply within one business day
Vision - line 3 camera
Warehouse detector v4 - 640x640 - centre aisle
38ms
Inference
24/s
Frames read
1,284
Objects tracked
OlDevs is a full-stack technology studio in Vancouver, British Columbia that designs and builds IoT and device management systems.A typical build covers device identity and provisioning, telemetry ingestion and storage, dashboards and alerting, over-the-air firmware updates, and integration with the business systems that act on the data. We choose connectivity such as cellular, LoRaWAN, Wi-Fi or BLE based on power budget, range and data volume rather than habit. Where an off the shelf platform already fits, we say so and help you configure it instead of building from scratch.
Key facts
Connected devices, honestly scoped
Every device gets a unique identity before it ships: certificates or keys provisioned at manufacture, a claiming flow that ties hardware to an owner, and a register that tracks state from first boot through to decommissioning.
Telemetry arrives in bursts, out of order and sometimes weeks late. We build ingestion that buffers, deduplicates and timestamps honestly, then store readings in a shape you can query later without paying to keep raw data forever.
Operators need to see fleet health in seconds and be told when something breaks. We build dashboards, maps and alert rules with sensible thresholds, escalation paths and enough context that nobody has to guess what an alarm means.
Firmware rollouts are where fleets get bricked. We build staged campaigns, signed images, version tracking, rollback paths and updates you can pause mid flight, so a bad build reaches a small test group rather than the whole fleet.
Not every reading belongs in the cloud. We push filtering, aggregation and local rules onto gateways or the device itself, so links stay cheap, latency stays low, and equipment keeps working when connectivity drops out.
Devices, data and dashboards each need their own protections: least privilege access, encrypted transport, key rotation and audit trails. Then we connect the platform to ERP, CRM, billing or maintenance systems so readings become action.
Process
Discovery
We map the devices, the readings that matter, who acts on them and what the platform must talk to. This is also where we test whether an off the shelf product would serve you better.
Architecture and pilot
We pick connectivity, data model, storage and hosting, then prove one device end to end. A pilot that survives a week of real conditions is worth more than a specification.
Build
Provisioning, ingestion, dashboards, alerting and update campaigns come together in weekly increments. You see a working demo every week and can redirect before anything hardens.
Field trial
A limited fleet runs in the real environment, with poor signal, power cuts and rough handling included. We tune thresholds, retry behaviour and update rings against what actually happens.
Launch and lifecycle care
We roll out to the full fleet, hand over documentation, runbooks and credentials, and stay on for monitoring, firmware maintenance and future work if you want us there.
How we work
Each unit carries its own credential, so the platform knows which device is reporting, what it is allowed to do, and whether it should still be trusted.
Cellular, Wi-Fi, LoRaWAN, BLE or Ethernet, picked for power draw, range and message size. Messages are small, queued locally and retried when the link returns.
Incoming payloads are validated, deduplicated and converted into one consistent schema, then written to storage tuned for time series queries and long retention.
Thresholds, absence of data and pattern rules generate alerts with severity and routing, so the right person is paged and everyone else is left alone.
Signed firmware moves through a canary group, then wider rings, with health checks between stages and an automatic stop if failures cross a limit.
Readings and events flow into maintenance, billing, ERP or CRM through APIs and webhooks, so a fault raises a work order instead of an email nobody reads.
Who it's for
Large fleets across sites and time zones, where the hard parts are role based access, integration with existing ERP and maintenance systems, and updating thousands of units without downtime.
Public assets, facilities and environmental sensing, where procurement, accessibility, bilingual EN and FR interfaces, data residency and clear audit trails matter as much as the readings themselves.
Identical equipment in many locations, monitored centrally. Head office needs comparable data across sites, while each operator needs a simple view of the few things they can actually fix.
Hardware products that need a platform behind them before launch. We start with the smallest system that proves the product, and keep the option to swap in a bought platform if the numbers favour it.
Weekly
Working demo cadence
1 business day
Reply to every enquiry
2014
Building software since
FAQ
Buy first, if you can. If your devices speak a common protocol and your needs stop at dashboards, alerts and basic rules, an established IoT platform will usually cost less and reach production sooner. Building makes sense when your data model is unusual, when the platform is part of the product you sell, or when per device fees stop making sense at your fleet size.
It depends on power, range, data volume and where the devices live. Cellular suits mobile or scattered assets. LoRaWAN suits low power sensors sending small readings over long distances. BLE suits devices near a phone or gateway. Wi-Fi and Ethernet suit fixed indoor equipment. We usually prototype more than one option before committing to a direction.
We do not publish figures, but the honest picture is that lifecycle cost is dominated by things that are easy to forget at kickoff: connectivity subscriptions, storage that grows every month, certificate renewals, firmware maintenance and support for devices already in the field. We size all of it with you during discovery, then you can request a quote.
Unique credentials per device, encrypted transport, signed firmware, key and certificate rotation, least privilege access for people and services, and audit trails. We also plan the unhappy cases: revoking a compromised device, rotating keys across a live fleet, and shipping a fix without physical access. Security is designed in, not added at the end.
Usually yes. We integrate with existing controllers, PLCs, gateways and third party sensors, often through a gateway that translates whatever the equipment speaks into one consistent message format. Where firmware cannot be changed, we adapt at the edge instead. We will tell you early if a device cannot be brought in safely.
You see a working demo every week. Early demos are usually one device sending real readings to a real dashboard, because a single end to end path proves more than months of diagrams. From there we widen the fleet, add rules and update campaigns, and run a field trial before the platform carries anything that matters.
You do. Clients own all code, designs, accounts and IP, including infrastructure configuration, firmware source and device certificates. We hand over documentation and runbooks so another team can pick it up. Ongoing care is an option and never a condition, and we never hold accounts or credentials back to keep you with us.
Let’s connect
Tell us what you’re building. We’ll reply within one business day with next steps and a tailored quote — no obligation.