Mobile operators have been switching off 2G and 3G for years to free up spectrum. For a tracking company that means part of the installed base will stop reporting, and the notice does not always arrive in time.
Why it happens and why it is not negotiable
Spectrum is finite. Keeping a 2G network alive for a tiny fraction of traffic, when that same band could carry 4G or 5G, does not hold up. The calendar varies a lot by country and operator, but the direction is the same everywhere.
What makes this awkward for tracking is that 2G was perfect for it: wide coverage, low power, cheap modules and tiny frames. Nothing that comes after is as convenient across all of those at once.
First step: know exactly what you have
Before quoting anything you need the real inventory. Not the one in a spreadsheet from three years ago: the one reporting today.
- Exact model and regional variant of every installed device
- The radio technology it actually uses, not the one on the datasheet
- Carrier and SIM type for each asset
- Age and condition of the device
- Associated customer and contract type
A decent platform gives you this in a report. If you have to assemble it by hand, you already know what your next problem after this one will be.
Which technology to choose for replacement
There is no single answer: it depends on the asset, on coverage and on whether the device has its own power.
| Technology | When it makes sense | What to watch |
|---|---|---|
| LTE Cat 1 | Vehicles with permanent power | The natural 2G replacement; good coverage and affordable modules |
| LTE Cat 1 bis | Same case, with lower module cost | Single antenna; slightly weaker at the edge of coverage |
| LTE-M (Cat M1) | Battery assets and light telemetry | Very low power and good penetration; uneven coverage by country |
| NB-IoT | Fixed or near-fixed sensors | Poor with mobility: it does not hand over between cells while moving |
| Satellite | Areas with no cellular coverage | High cost per message; used as a complement, not as the base |
A common mistake is buying NB-IoT for vehicles because the module is cheap and power draw is low. NB-IoT was not designed to move: an asset constantly changing cells loses reports.
How to plan the migration without stopping the operation
1. Classify by risk, not by age
Not every 2G device is equally urgent. Prioritise by the impact of going silent: a truck with valuable cargo under 24/7 monitoring cannot wait; a trailer reporting twice a day can.
2. Make the visit count
If you are already sending a technician to the vehicle, do not just swap the box. Check the harness, the sensor installation and the antenna placement. The visit is the big cost; the device is the small one.
3. Turn the problem into a commercial conversation
Replacement forces a conversation with every customer. It is the best opportunity in years to renew contracts, move someone to a richer plan or install devices that support what the old one could not: CAN, camera, fuel probe, Bluetooth sensors.
4. Do not migrate everything to one model
Concentrating the whole fleet on a single model leaves you exposed to a firmware problem or a manufacturer stock-out. Two equivalent models from different suppliers is a cheap policy.
What to ask of your platform during the migration
- That the new device keeps the asset's history rather than starting a blank one
- A report of assets by radio technology to track progress
- Automatic alerting for devices that stopped reporting
- Fast certification of whichever replacement model you choose
- That you are not billed twice during the device overlap
What not to do
Wait for the operator's notice. When it arrives the window is usually months, and the whole market goes out to buy the same module at the same time. Prices rise, lead times stretch and installers saturate. Migrating calmly costs considerably less than migrating against a deadline.
