Somebody in the office looks at the map at four in the afternoon and sees that a machine's latest reported position is from eleven that morning. They ring the site expecting a breakdown, and the site tells them it has been working all afternoon on the lower bench. Everyone involved now distrusts the system slightly more than they did that morning, and the actual cause was that the lower bench sits below the rim and there is no signal down there.
Coverage is not a footnote in this industry the way it is in road transport. The places heavy machines work are frequently the places networks were never built for: benches below the rim of a pit, deep cuts, tunnels, the inside of a shed, or simply the wrong side of forty tonnes of steel. A tracker designed around the assumption of a delivery van moving between towns on a main road will spend a great deal of its life in this industry doing something its designers never had to think about.
Two separate questions get conflated here, and separating them is most of the answer. The first is whether a unit can report at this instant. The second is whether the reading is lost. Only the first of these is genuinely about coverage, and it is the less important of the two, because almost nothing in asset management actually needs a reading within the second it was taken.
The SolidTrack Field Unit records on board when it cannot reach the network and sends what it held once coverage returns. The store-and-forward log holds three hundred entries by default. The practical effect is that the timeline usually fills itself back in. The afternoon on the lower bench is not lost from the record, it simply arrived late, and once it lands the recorded activity, the movement and the firing bursts all sit in their proper places. What you lost was the current picture on the map for a few hours, not the history behind it.
That distinction should change how you read the map. A quiet asset is a question, not a fault, and in a quarry it is usually a geographic question with a boring answer. Before anyone drives out to look at a machine that has gone silent, it is worth knowing which parts of your site do this routinely, because after a fortnight of operation the pattern is obvious and it is nearly always the same three places. A gap on the map is usually not a gap in the record.
Position deserves its own paragraph, because it behaves differently from data. GNSS supplies a reported position where conditions permit, and where the modem and the network support it, mobile cellular positioning can supplement that. Those two are not the same quality of answer. Cellular positioning is coarser: it puts an asset in roughly the right place rather than exactly the right place, which is entirely adequate for knowing which site something is on and not adequate for finding a specific machine in a large yard. It is also network-dependent, and it is not guaranteed to produce a fix at all. Availability, accuracy and freshness all depend on conditions, coverage and the configured reporting schedule, so it is worth knowing which of the two you are looking at, and how old it is, before you make a decision that depends on precision.
This has a direct consequence for alerting that people usually discover the hard way. An offline threshold set to the value that makes sense for a road fleet will fire constantly in a pit, and a rule that fires constantly gets muted, and a muted rule is the same as no rule at the moment you actually needed it. The threshold that flags a machine as offline should come from your own coverage pattern, once you have a few weeks of it, rather than from a value borrowed from somewhere else, and it will not be the same at every site.
The honest limits are worth stating. On board storage is finite: a machine out of contact for a very long time, at a high reporting rate, can exceed the storage it has, and the record from a period like that should not be assumed complete. And because the unit runs on its own sealed battery rather than the machine's, its life is counted in transmissions rather than in years, typically around four thousand of them at one transmission a day, with signal conditions and temperature both moving that figure. It is a transmission budget rather than a guaranteed lifespan. Reporting frequency is therefore a real trade off rather than a free setting, and a poor signal costs more than a good one because the unit works harder to get the message out.
The useful question to put to any supplier in this sector is not whether they have coverage at your sites, because nobody can promise that and the honest answer is that parts of your sites have none. It is what the unit does when there is none. Store and forward, a stated default log size, and a clear account of how position degrades when the sky is not visible is a system built by people who have been down a pit. A confident yes is usually a system built for something else.