The right link is decided by the site, not by the datasheet.

A decision that is expensive to reverse

Connectivity is often chosen early, quickly, and on the basis of what the team used last time. It then determines the power budget, the antenna and enclosure, the certification path, the backend architecture and the running cost per device for the life of the product. Changing it after deployment usually means new hardware.

It deserves more analysis than it typically gets.

What the choice actually turns on

• Coverage at the installation site. Not national coverage maps. Inside a steel building, in a basement, under a manhole cover or in a rural field, the answer is frequently different from the map.
• Power budget. The radio is usually the largest consumer on a battery device. Transmission frequency, payload size and protocol overhead often matter more than the technology’s headline figure.
• Data volume and frequency. A few bytes an hour and a continuous stream are different problems with different answers.
• Latency and direction. Whether the device must receive commands, and how quickly, rules some options out immediately.
• Running cost per device. Multiplied by the fleet and by the service life, this often exceeds the hardware cost difference several times over.
• Who controls the infrastructure. A private gateway you own behaves differently in procurement, in outages and in security review than a public network.
• Longevity. Network generations are retired. A product expected to run for a decade should not depend on a technology being switched off within it.

The options in practice

• Cellular. Wide coverage and no infrastructure to build, at the cost of higher power and a recurring subscription per device.
• NB-IoT and LTE-M. Lower power cellular variants suited to small, infrequent payloads, where the operator supports them in your region.
• LoRaWAN. Long range and very low power for small payloads, on public networks or gateways you own. Not suited to large or frequent data.
• Wi-Fi. Convenient where mains power and existing infrastructure are present, but dependent on credentials and on a network someone else manages.
• Bluetooth Low Energy. Excellent for short range and phone interaction, including commissioning, and often used alongside another link rather than instead of one.
• Wired. Ethernet, serial or powerline, which remain the most reliable choice wherever a cable can reasonably be run.

Many real products combine two: a low-power link for routine reporting and Bluetooth for local configuration and diagnostics.

How we help

• Requirements analysis. Turning the product’s behaviour into the numbers that decide this: payload size, reporting interval, acceptable latency, expected fleet size and service life.
• Site assessment. Establishing what coverage genuinely exists where the devices will be installed, by measurement where it matters.
• Power modelling. Estimating and then measuring the energy cost per transmission, and what it means for battery life. See battery and power engineering.
• Total cost comparison. Hardware, certification and recurring cost across the fleet over its life, not per unit in isolation.
• RF and antenna design. Antenna selection, placement and matching, and the enclosure interactions that so often undo a good radio choice.
• Implementation. Firmware, provisioning, reconnection behaviour and the backend that receives the data. See IoT development.

The failure mode to avoid

Choosing a technology, designing the hardware around it and discovering at installation that coverage is not there. A short, early feasibility exercise at real candidate sites costs a fraction of a hardware respin.

What you receive

A documented comparison of the realistic options against your requirements, coverage findings where assessment was performed, a power and cost model, and a recommendation with its reasoning recorded so the decision can be revisited if circumstances change.

Related

IoT development, legacy machine retrofit, EMC and compliance readiness.