Prismax: An IoT Data Platform That Runs on Radio, Not on Cell Coverage

The author works at PT Mitra Akses Globalindo (MAGNET), which develops Prismax. Coverage and savings figures below are vendor-stated and have not been independently verified by the author.

Remote industrial sites in Indonesia are large, flat and empty — and empty is what a cellular network does poorly. Plantations, mines and oil and gas fields spread over tens of thousands of hectares with tree canopy and spoil heaps between the crews, so a coverage map drawn for cities tells you nothing about planning telemetry across a block.

The problem is coverage, not bandwidth

Conventional telemetry assumes an IP path — LTE, GPRS, private LTE. That holds near settlements and roads, and thins out fast where the equipment is: a pump station mid-block, a haul truck on a bench, a wellhead with no line of sight to a mast. The failure is intermittent rather than clean: a device works at the gate, roams through a dead zone, and reports nothing for the whole window.

The approach: radio as the transport

Prismax uses Digital Mobile Radio (DMR) infrastructure as the data path instead of an IP bearer — a genuinely different engineering constraint, not a rebranding of mIoT on another bearer.

• DMR is two-slot TDMA: one 12.5 kHz carrier provides two timeslots, and data shares those slots with voice.
• It was designed for voice, so airtime is shaped by speech codecs rather than packet size.
• It is licensed and centrally coordinated, so it needs no SIM and no site backhaul to the internet — a field site can sit far from any fixed line.

Architecture

1. Prismax Node — attaches to the asset or its sensors, reads local inputs and GNSS, transmits over DMR.
2. Prismax Gateway — terminates the radio side, manages devices and buffered data, bridges to the central server.
3. Monitoring dashboard — real-time visualisation of position, telemetry and device health.

[field sensors] -> [Prismax Node] -> (DMR air interface) -> [Prismax Gateway] -> [central server] -> [monitoring dashboard]

Telemetry, GPS position, sensor data and SCADA commands all travel over the radio link without internet at the site. A control room can read and command assets where there is no WAN, because the network and the server are private to the operator.

Coverage figures: read these carefully

MAGNET has published two coverage figures, framed differently. Both are marketing claims, so neither belongs in a design unqualified.

• Up to 49 km per tower in dense-vegetation plantation areas, per MAGNET's PALMEX 2026 article — a linear reach for one tower under a stated land-cover condition.
• ±50 km radius, roughly ±100 km² from one tower, per MAGNET's February 2026 webinar with MYRZ — a radius figure converted into an area figure.

Three cautions before either number reaches a business case:

1. They are not the same claim. A 49 km reach and a ±50 km radius are the same order of magnitude in different units, and an area figure assumes a circle. Terrain is not a circle.
2. They are best case. "Up to" means an elevated antenna, clear line of sight, the right band. Treat "±100 km² from one tower" as a ceiling under favourable conditions, not a design assumption.
3. Real coverage depends on terrain, vegetation, antenna height and frequency band. Canopy attenuates and scatters the signal, and a valley floor will not see a ridge-top site well inside the radius. A site survey is mandatory, and the design should be validated against measured signal.

Security

• AES-256 encryption on the radio link.
• A closed-group / private network, keeping traffic separated from public networks so only provisioned radios join a group.

That is strong because the system is closed, not because radio is intrinsically unbreakable.

Store-and-forward

When the link drops, data is buffered locally and retransmitted automatically once connectivity returns. This is the difference between "we lost the data" and "we were slightly late".

On a radio network dropouts are a design input, not an exception — a device behind a ridge, or a mast losing power, will interrupt the flow. With store-and-forward the dashboard shows a gap that closes itself within buffer capacity; without it the same event leaves a silent hole in the record, and you have to assume the asset was healthy because nobody reported otherwise. Buffer depth is the real design parameter; size it for the outage you expect to survive.

Applications

Plantation

• Soil and weather condition monitoring across blocks.
• Fleet and heavy-equipment tracking, independent of cellular coverage.
• Harvest route optimisation for transport and loading teams.
• Automatic weighbridge recording, without manual entry.
• Panic button and man-down detection, working where LTE does not.

Mining

• Real-time asset and fleet tracking across pit and haul networks.
• Dynamic geofencing on equipment and personnel.
• Haul route optimisation, fuel and load monitoring.
• Underground safety monitoring for gas, temperature and humidity.
• Soil and irrigation monitoring, pipeline and wellhead monitoring.

The business case, stated honestly

MAGNET claims up to 15% fuel savings through idle-time monitoring and predictive maintenance. That is a vendor-stated figure, not something the author has independently measured, and the saving depends on how much of your fleet genuinely idles.

Complementary, not a replacement

As described at PALMEX 2026, MAGNET positions Prismax as complementary to existing infrastructure rather than a replacement — strengthening operations where other systems cannot reach, not displacing a working cellular or private LTE network.

When this is the wrong answer

• Dense urban coverage. Reliable LTE throughout means radio telemetry adds cost and complexity for no availability gain.
• High-bandwidth needs. Video streaming and large over-the-air firmware payloads are not what a narrowband voice network is for.
• Sites already fully covered. Where private LTE or fibre reaches the asset, use it.

Further reading

• MAGNET, PALMEX 2026 coverage article
• MAGNET and MYRZ Indonesia webinar, February 2026
• MAGNET company profile — https://www.magnet.id/tentang-kami/