Application · Modbus integration
Each remote node is a Modbus master for its own whole RS485 bus — one node, many devices — reporting over a private LoRa mesh on its own schedule, not on demand. A Siliqs gateway node listens to that mesh and keeps the latest reading from every device on every bus in a local cache. Your existing Modbus master polls the gateway exactly like any RS485 slave and gets an answer right away, in the same protocol it already speaks — no mesh round-trip, no new software.
A site has RS485/Modbus devices spread across places too far apart to share one wired bus — different buildings, different floors, opposite ends of a yard. Whatever CAN share a bus, does: one Siliqs node masters a whole run of RS485 devices at each location. It's only the distance BETWEEN locations that a wire can't cross — that gap is what the private LoRa mesh carries instead. That solves the reach problem. But the SCADA, PLC or historian that's supposed to consume this data was built for one thing: send a Modbus RTU request over a wire, get a reply back in milliseconds. It has no idea a mesh is even involved, and no patience for one.
One node listens to the mesh and holds a cache; your master keeps talking to it like any other slave.
One Siliqs node runs in gateway mode. It doesn't poll — it just listens to the mesh and keeps the latest reading for every device on every other node's RS485 bus, ready to hand out.
Your SCADA or PLC polls the gateway over RS485 exactly like any Modbus slave. It gets the cached reading back immediately — no waiting on the mesh, no change to your polling software.
Transparent: the gateway answers under each remote sensor's own original slave address — drop it in without touching your existing poll list. Aggregated: several sensors' readings sit under one shared slave address at different register offsets, when adding one device to your SCADA beats adding many.
Its limits are stated up front. What it won't quietly do is what makes it safe to poll.
No. It polls the gateway with ordinary Modbus RTU, over RS485, the same way it polls any slave today.
As fresh as the sensor's own mesh reporting interval. A configurable staleness threshold guards it — ask for a reading older than that and you get a Modbus exception, not stale data handed out silently.
Yes — that's transparent mode. The gateway answers under each remote sensor's own original address, so an existing poll list doesn't need editing.
Yes — aggregated mode. Several sensors' readings sit under one shared slave address, each at its own register offset, so your SCADA only needs to add one device.
No. This is a read path — the gateway reflects the mesh's latest telemetry, it doesn't relay writes or control commands.
Its cached reading keeps aging. Once it passes the staleness threshold, your master gets a clear Modbus exception instead of silence or a stale number — you find out, instead of quietly polling old data.
Tell us what your master already polls, how many sensors, and how fresh you need the data. We'll scope a gateway with you.
