Migration guide

Migrating from analogue or DMR to PoC, honestly

The honest starting point: not every fleet should migrate. If your people move between sites, drive, or work alone across a region, the move usually pays; if one well-covered site is your world, your DMR is probably fine. Here's how to decide — and how to run the switch without a single day of broken comms.

Should you migrate at all?

Migration is justified by geography, not by age of kit. The signals that a move to network radio will pay for itself:

And the honest counter-signals — where keeping DMR (or analogue) is the right engineering call: a single dense site whose direct radio coverage already works and costs nothing monthly; no-coverage environments — deep basements, tunnels, remote terrain with neither mobile signal nor Wi-Fi, where PoC physically cannot deliver live voice; and safety-critical loops that must survive a total network outage. Plenty of good migrations are partial: vehicles and supervisors to PoC, the factory floor stays DMR, a bridge joins them — and for many operations that's the correct end state.

The bridge pattern: nobody goes dark

The standard transition tool is a gateway: a unit that ties one conventional channel into one PoC talk group, so a site radio and a driver two counties away hear each other as if on the same system. This is what makes phased migration safe — the mobile team moves first, the site fleet keeps its familiar radios, and comms never break. Two practicalities from the field: place the gateway's donor radio where the conventional signal is strong (a windowsill in the site office, not a comms cupboard in the basement), and remember the bridged channel is still a licensed transmitter — the conventional side keeps its Ofcom licence during the transition, as covered in the PoC licence guide. Budget a few hundred pounds for the gateway — temporary scaffolding, or permanent in a hybrid fleet.

Talk groups are not channels

The commonest design mistake is recreating the old channel plan one-for-one: sixteen channels become sixteen talk groups, and the new system inherits the old system's habits and confusion. Talk groups are software — unlimited, instantly reconfigurable, and free — so design them around workflows, not frequencies:

Old thinking (channels)New thinking (talk groups)
Channel 1 = everyone, always congestedWorking groups per team/site, sized so most traffic is relevant to most listeners
Channel per site, no cross-site pathSite groups plus an operations group spanning sites, plus all-call for genuine broadcasts
Supervisors carry two radiosSupervisors scan multiple groups on one handset; dispatch sees everything
Emergency = shout over trafficPriority/emergency calls interrupt automatically, with GPS attached

Start with fewer groups than feels natural — three to six covers most 40-radio fleets — and let real usage argue for additions. Renaming and regrouping takes minutes from the platform console; on the old system it meant reprogramming every handset in the van.

Retraining: the half-day truth

Radio users need about half a day — plan for it honestly. The PTT button works exactly as it always did; what's new is selecting talk groups instead of turning a channel knob, the half-second latency (teach "press, beat, speak" on day one and it never becomes a complaint), and the features worth using deliberately — emergency button behaviour, who sees GPS and why (tracking lands far better explained than discovered), and messaging where it applies. Dispatchers and supervisors need a further half-day on the console. Run the training during the bridge phase, while the old radios still work — people experiment more freely when the safety net is visible.

Contract gotchas to catch before signing

The 90-day phased plan

PhaseWhat happensExit test
Days 0–30: pilot5–10 radios with the most mobile team; test the worst locations, vehicles at speed and Wi-Fi fallback per the pilot checklist; draft talk-group design; negotiate contract terms abovePilot log shows acceptable coverage; users would rather keep the new radios than give them back
Days 31–60: phased rolloutMigrate mobile teams and supervisors; install the gateway bridging old and new fleets; train everyone (half-day sessions); dispatch console liveA week of normal operations with zero comms escalations; old and new fleets talking through the bridge
Days 61–90: cutover and decide the tailMigrate remaining users or formally designate the retained DMR subset; decommission the bridge or make it permanent; cancel or retain the Ofcom licence to match what's left; sell surplus conventional kit while it holds valueNobody has pressed an old radio's PTT in a fortnight — or the hybrid boundary is documented and everyone knows which tool does what

At day 90, make the licence decision deliberately: if any conventional radio or gateway survives, it stays; if nothing does, let it lapse at renewal rather than paying £75 for nostalgia (details in the UK licensing guide).

The migration that fails is almost never a technology failure — it's the big-bang cutover where every radio changes on the same Monday, the basement turns out to have no Wi-Fi, and the old repeaters were switched off on Friday. Keep the old system alive until the new one has survived a real month.

FAQ

When is migrating from DMR to PoC worth it?

When geography outgrows one site: multiple locations, vehicles, regional lone workers, or a pending repeater purchase. A single well-covered site with working DMR usually shouldn't migrate.

Can old and new radios work together during the switch?

Yes — a gateway bridges a conventional channel into a PoC talk group, so both fleets share comms until cutover, or permanently in a hybrid design.

How long does retraining take?

Half a day for users, a further half-day for dispatchers. The PTT habit transfers directly; talk-group selection and the latency beat are the only new muscle memory.

Should we keep our Ofcom licence after migrating?

Only if any conventional transmitter — handset or gateway — remains in use. Fully migrated fleets can let it lapse at renewal.

Next: The full three-year cost comparison, itemised →