Who this recommendation is for
For the integrator and the network engineer who designs the connectivity of watches with the central system at the plant. The recommendation does not replace radio measurements in the field. It structures the moment of decision: when we stay with the simple solution and when we invest in a standard.
Recommendation 1: stay with simple LoRa while the scale is small
As long as the number of watches is small and the site has a single location, a simpler network is cheaper to commission and easier to maintain. Fewer components mean fewer points of failure and fewer things to configure at the client. Do not move to LoRaWAN just because it is a standard, if nothing requires it.
Decision threshold versus LoRaWAN gateway capacity
Recommendation 2: treat 200-300 watches as a signal for analysis
The threshold of about 200-300 watches is not a hard limit. It is the moment when it is worth calculating the network load and deciding on migration. Below the threshold you do not make changes. Above it you analyse the traffic, the number of alarm messages and the extent of the site. Such a signal should be written into the expansion plan already in the first offer, so that the client does not learn about it only after the expansion.
Recommendation 3: use the capacity of the LoRaWAN network
In the LoRaWAN option, an 8-channel grid and adaptive data rate (ADR) provide capacity on the order of a thousand watches per gateway. ADR selects the data rate to the radio conditions of each device, so watches close to the gateway do not occupy the link longer than necessary. For a site that is growing, this means spare capacity without adding further gateways only because of the number of devices.
Recommendation 4: require keys per device
In LoRaWAN the keys are per device. This is an important advantage for the IT and security departments: losing one watch does not force a change of a shared key for the whole fleet, and revoking one device does not affect the others. With a shared key, every incident involving one watch touches the entire network.
Recommendation 5: do not change the backend, add a bridge
With the LoRaWAN option the InnerWeb backend does not change its contract. A thin bridge decodes the data from the network server and passes it on as one more gateway. As a result the migration is a network task, not a rebuild of the application. Plan the bridge, the ChirpStack server and a test of one gateway with a group of watches before the full switchover.
- Pilot on one gateway and a group of watches.
- Parallel operation of the old and new network for the transition period.
- Switching the watches in groups, with the option to roll back.
Recommendation 6: decide who maintains the network server
ChirpStack is an additional component. Agree at the start who maintains it: the integrator, the client's IT department, or InnerWeb within the on-premise deployment in the plant network. Without this agreement the migration ends with no one responsible for updates and configuration backups.
Video
Frequently asked questions
Is LoRaWAN the default InnerWeb option?
No. It is an option on request, used at a larger scale or when the client requires a standard.
Does the application have to change during the migration?
No. The backend contract stays, and a thin bridge decodes the data from the network server and passes it on as one more gateway.
Why are keys per device important?
Because the compromise of one watch does not force a key change across the whole fleet. Revocation applies to one device only.
Let's talk about the beacon layout at your plant
We will show the map, the readings and the configuration on your site plan. Book a short conversation with the InnerWeb team.
Let's talk