Process
Hardware NPI process for LoRaWAN and IoT products. Seven steps, one team. Timelines are quoted per project — not as a catalog promise.
What each step actually produces
Schematic, PCB, and 3D-print tiles are public examples (Wikimedia Commons). Workshop photos show the floor — no generated box diagrams.
01
Every engagement starts with a mutual NDA — yours or ours. Then we listen: what the product does, who buys it, what it costs to build, and what "done" looks like. You talk with engineers who ask about constraints, not a sales team taking notes.
You leave this step with: Signed NDA and a written scope outline
02
We map the product into blocks: power, LoRaWAN or other radio, sensing, compute, enclosure. You get a written architecture — key components, battery budget, regional band plan, firmware outline, and a cost model you can defend — before anyone draws a schematic.
You leave this step with: Architecture note and cost model
03
Electronics, firmware, and industrial design move together. Antenna keep-out, battery, seals, and code paths are decided in one project, not four contracts.
You leave this step with: Schematics, layout, firmware draft, CAD
04
SMT a small board run. Print or CNC the housing. Prove radio, sensors, fit, and current draw against the architecture.
You leave this step with: Working prototypes and bring-up reports
05
Automated DFM checks plus an engineer review: panelization, test points, assembly clearances, mold draft. Free fixes stay free.
You leave this step with: DFM report and frozen file set
06
A short production pass with the fixtures and scripts that will run at volume. Yield and failures get a conversation, not a change-order chain.
You leave this step with: Pilot units, fixtures, yield notes
07
Volume under the same team that designed the product. You own the files. We support the line when something drifts.
You leave this step with: Production units and pack-out
Anonymous project notes — problem, what we did, what the partner kept. No customer names.
Problem. An outdoor environmental node needed IP-rated housing. The first ID concept boxed the antenna and killed range.
What we did. Electronics and industrial design sat in one review: antenna keep-out, gasket, and wall thickness. 3D print for feel, CNC for RF and fit, then mold-ready CAD.
What they kept. A housing that still meets the radio budget. The partner owns the CAD, keep-out drawing, and DFM notes.
Problem. A facility node needed extra I/O and a different battery than the reference. The team did not want to buy a steel tool on an unproven layout.
What we did. Schematic and layout with RF and battery isolation on pass one. Automated DFM, then a small SMT run and bring-up report.
What they kept. Working prototypes and a file set (Gerbers, BOM, pick-and-place) ready for pilot — without a second vendor re-learning the board.
Problem. The same end-node family had to join EU868, US915, and AS923 networks with one payload map.
What we did. Firmware variants on a shared radio stage: join, ADR, and documented uplink fields. Current draw measured against the battery budget.
What they kept. Band SKUs the partner flashes and points at their own network server. Payload notes stay with them.
Services, Process, and Why ParticlIO — then Contact Us so an engineer can read your brief.
IoT hardware design and manufacturing — PCB, firmware, enclosure, and assembly in one flow.
See servicesOne team for OEM/ODM LoRaWAN — design-house depth with manufacturing follow-through.
Why ParticlIO