Mixed-fleet support has to be checked by equipment make, model, software version and task, not by a single compatibility label. Reading telemetry, sending dispatch instructions and controlling equipment are three different capabilities, and a supplier that demonstrates one may not demonstrate the others. Ask for evidence of the exact workflow on the mine's equipment and network before treating a mixed fleet as supported.
That sounds obvious until a compatibility matrix arrives with one tick against a machine family. The tick usually means data left the machine during a demonstration. It does not say which fields or how often. It does not say which direction the data moves, over which network, on which software version, or who fixes the interface when it stops. This article sets out the boundaries to test and the records to ask for. The fleet management and dispatch buying guide covers the wider selection questions; this article stays on acceptance evidence. It draws on a dated audit of 25 fleet-management-dispatch listing descriptions in the MSR directory, and it states the limits of that audit plainly.
Define support at three different boundaries
A mixed fleet creates at least three integration problems, and they carry different risks. A supplier may be strong at one and weak at another.
Collecting telemetry
Telemetry is a read path. Data leaves the machine and arrives somewhere the mine can use it. The questions start with what data, at what interval, from which machine system.
A truck's OEM telematics feed may expose engine hours, fuel burn, payload and location while withholding the machine's internal control states. A third-party gateway may read the machine bus directly, which can reach more signals but requires hardware, installation and a data dictionary.
Ask which fields are available by make and model, whether they come from the OEM's own telemetry service or from a device fitted to the machine, and what happens when the OEM changes its software or its data plan. Ask where the data is stored and who can read it. A telemetry demonstration is evidence of collection. It says nothing about dispatch or control.
Delivering assignments and operator acknowledgements
Dispatch is a write path. An assignment has to reach the operator, the operator has to accept or reject it, and that response has to return to the dispatch system as a record. On a mixed fleet, the hard part is the last metre. The assignment may have to land on an OEM's in-cab screen, a mine-supplied tablet, a radio call or a paper board.
Ask for the exact path on each machine type. Does the assignment appear on the OEM's screen or on a separate device? Who confirms receipt? Is the acknowledgement timestamped and attributed to a named operator? What does the dispatcher see when a machine has not responded? A screen that displays an assignment is not the same as a system that records acceptance.
Equipment control is a separate safety scope
Controlling equipment, whether autonomous haulage, tele-remote operation or automated drilling, sits behind a safety case, certified hardware and interlocks. A data feed that reads machine state is not a control channel, and a dispatch integration does not confer permission to command a machine.
Keep the two procurement conversations apart. If a supplier offers both, ask for the control system's safety approvals and for the boundary between the data integration and the control path. A mixed-fleet data platform may have no control capability at all. That is a scope statement.
What 25 fleet listings explicitly mention
The MSR fleet-management-dispatch inventory contained 25 listings on 10 November 2026. Two Eclipse deployment listings remain separate rows. The table records what their descriptions say under four literal codes. A listing counts once per code, codes overlap, and counts are listings, not vendors or tested systems. The descriptions are vendor claims. They are also truncated in the source data, so silence is unknown rather than evidence of a missing function.
| Code | Inclusion rule (wording in the listing description) | Listings |
|---|---|---|
| O | Explicit any-OEM, non-Sandvik, OEM-agnostic or mixed-fleet wording | 2/25 |
| T | Literal telemetry wording | 2/25 |
| D | Explicit dispatch or assignment wording, including dispatch personnel and automated mission assignment | 9/25 |
| I | Explicit integration wording, an API, or SQL databases named as an integration path | 6/25 |
OEM scope, telemetry, assignments and interfaces
O, any-OEM or mixed-fleet wording (2/25). GroundHog FMS Open Pit says its system works "for any OEM equipment". OptiMine says it integrates data "including non-Sandvik equipment". Both are vendor claims. Neither description names the makes, models or software versions covered, and neither sets out which functions are supported across them.
T, literal telemetry wording (2/25). The Eclipse Miningware (On-Premise) and Eclipse Miningware (SaaS) descriptions both say the product brings "telemetry data" together with mobile plant, grade control, drill and blast and maintenance. They are separate listings of one deployment family, so the count is two listings.
The code counts the word as written. FleetOps describes GPS-based production capture without using "telemetry", so it is not counted. The count is not a census of telemetry capture across the category.
D, dispatch or assignment wording (9/25). AutoMine mentions "automated mission assignment". DISPATCH Fleet Management System automatically assigns haul trucks. FleetControl adds "active dispatch". GroundHog FMS Open Pit claims "AI-driven dispatch optimization". Hexagon Mission Manager is a "dispatch optimization module". Mine4D Production aids "dispatch personnel". MineStar Fleet is described as a dispatch and fleet management system that assigns trucks. Minetrack is an "electronic dispatch" system. Wencomine plans "truck dispatching priorities".
The code records wording. It does not establish that any of these systems can dispatch another maker's machines. A product name alone would not earn the code.
I, integration wording (6/25). GroundHog FMS Open Pit claims "open APIs". HxGN MineOperate UG Pro cites "standard SQL databases enabling integration with third-party" systems. Mine4D Guidance integrates with the Mine4D FMS. MineStar Command integrates with MineStar Fleet. OptiMine integrates data from assets and people. Proximiti integrates GPS collision awareness with radar.
I counts internal product and sensor integration as well as third-party interoperability, so it does not establish an open interface. Wencomine's truncated description stops at "SQL-compliant dat" without an integration claim, so it is not counted. If the full description made that claim, the count would change.
What missing wording cannot tell us
No description sets out a compatibility matrix by model and software version. None states interface direction and latency, the licensing terms that cover an interface, or who owns interface support. Four descriptions mention hardware: commercial tablets, an industrial tablet platform, on-board hardware and proprietary communication equipment. None lists the hardware required per machine type. Those absences are the reason to ask for documents. They are not proof of a gap.
The audit also cannot tell you about market coverage or product quality. It counted descriptions supplied on one date. Listings can change without notice, and a longer description may earn a code that a shorter one does not. Use the table to frame questions for suppliers. It is not a shortlist.
Build the mine's compatibility register
Before the first workshop, build a register with one row per machine family and one column per question. Keep it in a form the mine controls, with a date and an owner for each answer.
Make, model, version, required hardware and interface owner
Record the make, the model and the software or firmware version in service. Then record the OEM system the machine reports to, the fields the mine needs, the protocol or interface, and the direction (read, write or both). Add the hardware required at the machine and in the field, who installs and maintains it, the licence that covers it, and who owns the interface on each side.
Two columns do most of the work. Label them "evidence seen" and "date tested". A vendor datasheet belongs in the first. A signed test record belongs in the second. If an interface relies on an OEM service, note the OEM's support window and the notice period for changes.
Surface and underground connectivity constraints
Surface fleets usually assume broad coverage across pits, dumps and haul roads. Underground fleets deal with headings, declines, corners and moving faces, where coverage is a design problem rather than a given. The same compatibility question has different answers in each setting.
Ask where the integration runs: at the machine, at a gateway, at a site server or in a vendor cloud. Ask what the system does when the link drops, how long it buffers data, and whether dispatch can still issue assignments offline. A store-and-forward design may preserve records without preserving real-time dispatch. Test the behaviour the mine actually needs. Our article on underground connectivity and monitoring platforms covers how to test the network underneath.
Deployment boundaries matter for the same reason. On 11 September 2026, International Mining reported on Nokia's Cognitive Operations platform. The report described a mining product available for on-premises deployment or through the Microsoft Azure Marketplace, with AI processing at a field edge node. Whatever the platform, a mine that mixes on-site and cloud components needs to know which data crosses which boundary, who can read it, and what still works when the link to the cloud is cut. Nokia's announcement is a prompt for those questions. It is not evidence about fleet interoperability or dispatch capability.
Demonstrate a complete dispatch round trip
Documentation describes a path. A round trip proves it. Choose one truck, one loader and one operator, and run the workflow the mine uses on a normal shift.
Timestamped event, assignment, acknowledgement and reconciliation
Create an assignment in the dispatch system and record the time. Record when it appears on the machine, when the operator accepts it, when work starts and when the load is completed. Compare those timestamps with the records in the OEM system and in the mine's production report.
Ask for the reconciliation method. If the two systems disagree, which record is authoritative, and who resolves the difference? A demonstration that shows data moving is useful. A reconciliation that matches the shift report is acceptance evidence.
Disconnection, duplicate event and recovery
Break the link on purpose. Pull the network at a safe point, let the machine keep working, then restore it. Check what arrives, in what order, and whether anything is duplicated or lost. Repeat an event, or replay a message, and see whether the system recognises it as the same event.
Then change something small: a software update, a new machine, a renamed location. Watch whether the interface survives. Record who fixed it and how long it took. These tests cost a shift. They are cheaper than finding the gap after go-live.
Record exclusions and ongoing support responsibilities
Write down what the demonstration did not cover: machine models outside the test, functions that stayed read-only, sites with different network conditions, and software versions not tested. An exclusion list is a useful document. It tells the next shift what to watch.
Put the support model in the contract. Who is the first call when an interface breaks, and what response time applies? Who tests the interface after an OEM updates its software or a vendor changes its API? How much notice will the mine get? What happens to the data and the interfaces if the contract ends?
A compatibility label is where the mixed-fleet question starts. A register the mine owns, a round trip it has witnessed, and a support agreement that names the person who fixes the link are what close it.
Sources: MSR fleet-management-dispatch listing descriptions supplied on 10 November 2026 (25 listings, vendor-written and truncated in the source data). International Mining, "Nokia launches AI and edge platform to help run safer, more resilient field operations", 11 September 2026 (read 10 November 2026). Product names link to their MSR profiles. Listing descriptions and the Nokia deployment details are vendor and publisher claims. MSR has not tested these systems.
This article is general guidance, not engineering, procurement or safety advice. Test any integration on your own equipment, software versions and network, with your own operations and safety teams.
Related Categories
Explore the software categories referenced in this article.
Related Buying Guides
Our guides to what to check before you shortlist software in these categories.