Straight answers to the questions operators actually ask — no product names, no sales pitch. Where the honest answer is “it depends”, that is what it says, along with what it depends on.
The foundations — what it is, what it collects, what it returns and where the law sits.
Telematics is the combination of telecommunications and data. In practice it means a small device fitted to a vehicle that collects information — where it is, how fast it is going, how it is being driven, what the engine is doing — and sends it over the mobile network to software you can look at on a computer or phone.
The plain version: it is a way of knowing what your vehicles are doing without having to phone someone and ask.
Vehicle tracking is the location part of telematics — a dot on a map showing where each vehicle is and where it has been.
Telematics is the broader category. Modern telematics includes tracking, but adds driver behaviour, engine diagnostics, fuel or energy use, idling, mileage, servicing data, walkaround checks and much more. Most suppliers use the words interchangeably; what matters when buying is what the system actually does, not which word is on the brochure.
Four steps, and none of them are complicated:
Typically:
What it does not collect, unless you specifically add it: audio, video, or anything about the driver as a person beyond how the vehicle was operated.
For most commercial fleets, yes — and usually faster than expected. But the answer deserves more honesty than a headline percentage.
The returns come from a small number of reliable sources:
The critical point: the technology does not deliver the savings — the changes you make with the data do. A fleet that installs devices and never reviews the reports gets nothing. A fleet that reviews weekly, coaches monthly and adjusts routes quarterly gets a great deal. Any supplier who tells you the savings are automatic is selling you a disappointment.
In rough order:
Yes, when it is done properly. Tracking vehicles used for work is a well-established, lawful business activity in the UK and across Europe.
The legal obligations sit on how you do it:
None of this is onerous. All of it is much easier to do at the start than to retrofit after a grievance.
Generally, no — and relying on consent is usually the wrong approach. Regulators take the view that employees cannot give genuinely free consent to their employer, because of the power imbalance. If you rely on consent and one driver withdraws it, you may have to stop monitoring that individual while continuing with everyone else — an unworkable position.
The correct approach is to identify a proper lawful basis (usually legitimate interests), document the assessment, and be transparent.
Consent is not the same as consultation, however. You should absolutely consult your drivers and their representatives before you begin. Not because the law demands their permission, but because it produces a better system, fewer grievances and far better results.
Only in narrow circumstances, and generally you should not. Where private use is permitted, the expectation is that drivers can switch to a privacy or private mode which stops recording detailed location, while the vehicle remains protected for theft purposes.
If a system you are evaluating cannot do this and your drivers take vehicles home, that is a genuine compliance problem, not a minor feature gap.
Position accuracy is typically within a few metres in open conditions, degrading in urban canyons, dense woodland, tunnels and multi-storey car parks. Modern devices use multiple satellite systems simultaneously, which materially improves performance in cities.
Two things affect accuracy more than the satellites do:
Positioning works globally. The limiting factor is mobile network coverage for sending the data. In a coverage gap, the device stores data and uploads it when signal returns — so nothing is lost, but live visibility is interrupted.
For rural operations and cross-border work, ask specifically about roaming or multi-network SIMs, which allow the device to use whichever network is strongest rather than being locked to one.
A properly installed system draws negligible power and does not affect vehicle performance. Reputable installers work to manufacturer guidelines, and installation by a competent installer does not invalidate a vehicle warranty. Poor installation can — which is an argument for using qualified installers rather than for avoiding the technology.
Parasitic battery drain is a legitimate question for vehicles that stand for long periods. Well-designed devices go into a low-power sleep mode; ask what the standby current draw actually is and at what voltage the device stops drawing power to protect the battery.
A plug-in diagnostic-port device takes minutes and requires no downtime. A hardwired device typically takes 45 minutes to two hours per vehicle. A multi-camera installation on a large vehicle takes considerably longer — half a day is common.
The disruption is almost never the installation itself; it is the scheduling. Fleets that plan installation around existing service visits, MOT dates or shift patterns barely notice it. Fleets that try to install everything in one week discover their vehicles are all needed that week.
No. Small fleets often see proportionally larger benefits, because they typically have no dedicated fleet manager and therefore more untapped inefficiency. A five-vehicle operation losing an hour a day to idling and unnecessary mileage is losing a meaningful share of its margin.
What differs is the approach. Larger fleets need integration, role-based access, structured reporting and formal coaching programmes. Smaller ones need something that works out of the box and does not require someone to become a part-time data analyst.
Common and fair. Two questions usually settle it:
Changing supplier is not automatically the answer. Sometimes the existing system is capable of far more than is being used, and the real gap is process, configuration and training.
Telematics becomes more useful, not less. EV-specific data includes state of charge, remaining range, charging sessions and locations, energy consumption per mile, regenerative braking use and battery health indicators.
The operational questions change too: not “where is the nearest fuel card site” but “will this vehicle complete its round, and where does it charge if not”. Fleets running mixed diesel and electric usually find they need a single platform reporting both consistently, or comparison becomes impossible.
Cameras, evidence, driver acceptance and what the footage is actually worth.
Video telematics is vehicle cameras combined with vehicle data, so that footage arrives with context. Instead of a video clip you have to hunt for, you get an event: what happened, when, where, at what speed, with what braking and steering, and who was driving — with the footage attached.
The distinction matters. Video on its own tells you what a camera saw. Video telematics tells you what happened.
A consumer dashcam records to a memory card in the vehicle. To see anything, someone has to physically retrieve the card. If the card has failed — and consumer cards fail regularly in the heat, vibration and constant rewriting of a working vehicle — there is nothing on it, and nobody knew.
A video telematics system is connected. Footage and data travel over the mobile network. Incidents are detected automatically and flagged immediately. The system monitors its own health and tells you when a camera or card has failed. Footage can be requested remotely, and access is controlled and logged.
The practical difference: with a dashcam you find out whether you have evidence after you need it. With video telematics you know before.
An MDR — mobile digital recorder — is a ruggedised multi-camera recorder built for commercial vehicles. Where a dashcam handles one or two views, an MDR handles four, eight or more, recorded in sync.
You need one when a single forward view does not cover where your incidents actually happen. On an articulated vehicle, a refuse collection vehicle, a tipper, a coach or a vehicle with a tail-lift, the majority of incidents occur somewhere other than directly ahead — reversing, manoeuvring, at the kerbside, or during loading. Forward-only coverage on those vehicles records the one direction where least goes wrong.
Stripped of marketing language, three things:
The third one is the least glamorous and the most commercially important. The value of AI in this context is mostly in what it stops sending you.
No, and it is a decision to make deliberately rather than by default.
The case for: the single largest contributory factor in commercial vehicle collisions is what the driver was doing in the seconds beforehand. A forward camera shows a collision happening; a driver-facing camera shows whether distraction or fatigue caused it — and, just as often, proves it did not. In a serious injury claim, being able to demonstrate the driver was attentive and belted is a powerful defence.
The case against: it is the most intrusive element of any programme, generates the most resistance, requires the most careful legal and policy work, and if handled badly can damage trust across the whole workforce.
The practical middle ground many operators adopt: fit driver-facing capability, but configure it to record only around triggered events rather than continuously; process the analysis on the device rather than streaming it; disable audio by default; make the retention period short and automatic; write the policy with driver representatives before installation; and be absolutely explicit that footage is used for coaching and defence, not for surveillance of breaks and conversations.
Yes, subject to doing it properly. In-vehicle cameras are widely and lawfully used. The obligations are the same as for telematics but the bar is higher because the intrusion is greater:
This is the question that determines whether the investment works, and it is answered by process rather than technology.
What consistently works:
That is your decision to make and document. A defensible arrangement typically includes: named individuals with defined access rights; an audit log recording every footage view and download; automatic deletion after a stated period unless the material is flagged for a claim or investigation; and a formal process for releasing footage to insurers, police or claimants.
Continuous footage held on the vehicle usually loops over a period of days to weeks, depending on storage size, number of cameras and quality settings. Footage uploaded because of an event is retained in the platform for a defined period — commonly 30 to 90 days as a default, extended where a claim is open.
Keeping everything indefinitely is expensive and hard to justify. Set a retention period, write down the reason, and automate the deletion.
Generally yes, provided its integrity can be demonstrated. What matters is a clear, unbroken chain of custody: footage that can be shown to be original and unaltered, with a reliable timestamp and location, an auditable record of who accessed it, and no gaps in the sequence. Systems that download footage to a laptop and email it around are weaker on all four counts than systems that retain the original and control access centrally.
More than people expect if the system is badly configured, and very little if it is well configured. Continuous streaming of high-definition video from every vehicle would be prohibitively expensive — which is why sensible systems do not do it. They process events on the device, upload short clips, and stream live only on request.
Questions worth asking: what is the monthly data allowance per device; what happens if it is exceeded; how much data does a typical footage request consume; and is the allowance pooled across the fleet or fixed per vehicle.
Modern vehicle cameras use infrared or low-light-optimised sensors for interior views and high-dynamic-range imaging for exterior views, which handles the hard cases — emerging from a tunnel into bright sun, or a dark road with oncoming headlights. Performance still degrades in heavy rain, spray, fog and on a dirty lens.
Two practical points that matter more than sensor specifications: lens cleaning needs to be part of the daily walkaround routine, and camera positioning within the windscreen wiper sweep is not a detail — it is the difference between usable and useless footage in winter.
It will, eventually. Vehicle cameras operate in one of the harshest environments consumer electronics ever encounters: constant vibration, temperature extremes and continuous rewriting of storage.
The important question is not whether components fail but whether you find out. A connected system with automated health monitoring reports a failed camera, a failed card or a disconnected device the same day. An unconnected one reports it the moment you need footage and discover there is none — which is invariably the day of your most serious incident.
Ask any prospective supplier how device health is monitored, how failures are reported, and what the contracted response time is.
The financial case rests on four pillars, in rough order of speed:
Set against that: hardware, installation, subscription, data, and — the cost most often forgotten — the staff time to review events and run coaching. A programme with nobody assigned to run it delivers pillar one and none of the others.
Equipment, plant and everything that does not drive itself.
Asset tracking is knowing where your equipment is, whether it is being used, and being told if it moves when it should not.
It applies to everything that is valuable but does not drive itself: trailers, excavators, dumpers, generators, compressors, welfare units, site cabins, traffic management equipment, skips, cages, containers, pumps, access platforms, medical equipment, and — at the smaller end — tools and instruments.
Three practical differences change everything about the technology.
Power. A vehicle supplies power continuously. Most assets do not, so the tracker runs on its own battery and everything is designed around conserving it.
Movement pattern. A vehicle moves most days on predictable routes. An asset may sit still for six weeks and then move once — and that single unexpected movement is the whole point of the system.
What you need to know. For a vehicle you want continuous detail. For an asset you usually want two things: is it where it should be, and is it being used. Reporting twice a day may be entirely sufficient — and lets a battery last for years instead of weeks.
Powered assets — excavators, generators, telehandlers, refrigeration units — have their own power source. A tracker can be wired in and report frequently, and can also read genuine machine data: engine hours, fuel level, fault codes, and in many cases operating mode. This gives real utilisation data rather than “it moved” data.
Non-powered assets — trailers, skips, cabins, cages, barriers, containers — need a self-contained battery device. Design priorities shift to battery life, ruggedness, weatherproofing, concealed mounting and tamper detection.
Many operations need both, which is a good reason to want one platform that handles both rather than two systems and two logins.
Anywhere from several months to several years, driven almost entirely by how often the device reports.
A device reporting once a day may run for five years or more. The same device reporting every two minutes may last weeks. This is the single most important specification conversation to have, and the one most often skipped: decide what you actually need to know and how quickly, then set the reporting strategy accordingly.
Sophisticated devices vary their behaviour — sleeping when still, waking and reporting frequently when moving. That gets you long battery life and useful detail during a theft, which is exactly when you need it.
Rather than value alone, four tests are more useful:
Value alone is a poor test. A £900 item that stops a £40,000-a-day site is worth tracking; a £15,000 item that sits in a yard and has three spares may not be.
It improves the odds substantially, particularly in the first hours after a theft. Recovery rates for untracked plant and equipment are low across the industry — most of it is never seen again, and once identifying marks are removed, tracing becomes almost impossible.
Tracking changes the economics in three ways: it alerts you to unauthorised movement immediately rather than when someone next visits the site, often at night when the theft is happening; it gives police an actual location rather than a description; and its visible presence acts as a deterrent.
Realistic expectations matter. Determined thieves look for trackers, and organised groups may use signal jammers or strip components rather than take whole machines. This is why the practical answers are concealed mounting, more than one device on very high-value assets, tamper alerts and — critically — a written plan for what happens at 2am when an alert fires. An alert nobody acts on for nine hours is not a recovery capability.
For most operations, theft is the reason they buy it and utilisation is where the money actually is.
Yes, using low-power Bluetooth tags rather than satellite trackers. A tag is small, cheap and lasts a long time, but it does not know where it is — it is detected by a nearby reader, typically a van-mounted gateway, a site beacon or a phone.
That gives you “last seen at this location at this time” rather than live location. For a toolbox, that is usually exactly what you need: which van it is in, whether it came back from site, and who had it last. For a £40,000 machine, it is not enough.
The economics are the point. Tagging an entire tool pool becomes affordable at a per-item cost where satellite tracking would not be.
Satellite positioning needs sky visibility, so a device fully enclosed in a metal container or deep underground will struggle to fix a position. Sensible systems handle this by reporting the last known good position with a timestamp, by falling back to approximate positioning from mobile network masts, or by recording the sequence of events so you can see where an asset was when it was last visible.
Low-power network technologies designed for these devices penetrate buildings considerably better than standard mobile signal, which helps for the transmission even when positioning is degraded. For genuinely enclosed environments, short-range tags with fixed readers at entry points are usually the better architecture.
Costs are driven by device type and reporting frequency more than by anything else. A long-life battery tracker reporting daily is inexpensive to buy and inexpensive to run. A hardwired device on a machine, reporting frequently and reading machine data, costs more on both counts. Bluetooth tags are cheaper still per item but require gateway infrastructure.
The number to build a business case on is not the per-device cost. It is the annual cost across the whole population set against a realistic estimate of losses avoided, hire spend reduced, equipment not purchased, and time recovered.
Yes, and it is one of the more valuable applications — but it requires agreement. Fitting a tracker to equipment you do not own needs the owner’s permission, and tracking equipment operated by another company’s staff raises the same data protection considerations as tracking your own.
Many hire companies now provide tracking data as part of the hire, which removes the problem entirely. Where you are the hire company, the ability to give customers visibility of what they have on hire is increasingly a competitive requirement rather than a differentiator.
The practical questions that decide whether the investment actually works.
Start with whichever addresses the problem that is currently costing you most. In practice:
The one thing worth doing regardless of where you start: check that whatever you buy first can extend to the others without replacing it. Buying three separate systems from three suppliers over four years is the most common and most expensive path through this market.
Increasingly, yes — and there are good reasons to want it: one login, one support route, one contract, one set of reports, and the ability to compare across the whole operation. Where different data sits in different systems, someone ends up reconciling it in a spreadsheet, and that person becomes a single point of failure.
Two honest caveats. First, a single platform covering everything adequately may be a worse outcome than two platforms each covering their area excellently — the right answer depends on where your risk actually sits. Second, “one platform” sometimes means several acquired products behind one login, which is not the same thing. Ask to see a single report combining data from vehicles, cameras and assets. The answer is usually informative.
Pricing is normally structured as hardware plus installation, then a per-device monthly or annual subscription covering software, connectivity and support. Video systems cost more than basic tracking, principally because of hardware, storage and data.
The number that matters is total cost of ownership across the contract term, which should include:
Buying outright means a capital cost and asset ownership, usually with a lower ongoing subscription. Including it in the subscription means no upfront capital, a predictable operating cost, and typically hardware replacement covered within the term.
Beyond accounting preference, the practical questions are: what happens if a device fails in year four; who pays to move devices between vehicles; and what happens to the hardware if you change supplier. Bundled hardware is often effectively non-transferable, which is a switching cost worth pricing into the decision at the outset.
Three to five years is normal, because hardware and installation costs are amortised across the term. Shorter terms are available at higher monthly rates.
More important than the length is what happens inside it: whether pricing is fixed or indexed, whether you can add and remove vehicles freely, what the notice period is, whether it auto-renews, and — the one to check carefully — what happens to your historical data when the contract ends. Data portability is easy to agree at the start and impossible to negotiate at the end.
You should. Make sure the contract says so explicitly, and that it also covers: your right to export your data in a usable format at any time; what the supplier may do with it, including whether it is aggregated or anonymised for other purposes; how long they retain it after the contract ends; and where in the world it is stored and processed.
For a straightforward tracking deployment on a modest fleet, a few weeks from order to fully operational. For a multi-camera installation across a large or complex fleet, several months — driven by installation scheduling and vehicle availability, not by the technology.
The realistic constraint is almost always the same: getting vehicles to installers without disrupting operations. Fleets that plan around existing service schedules do this without pain. Fleets that do not, discover their vehicles are all needed at once.
More people than most organisations initially expect, and the omissions are predictable:
Rarely the technology. Nearly always one of these:
Capture a baseline before installation. This takes an afternoon and is impossible to do retrospectively. As a minimum, record:
Then review at three, six and twelve months against those same measures. Programmes that do this get renewed easily and expand. Programmes that do not spend their renewal arguing about anecdotes.
A connected device on your vehicles is part of your attack surface, and should be reviewed accordingly. Reasonable questions: how is data encrypted in transit and at rest; how are devices authenticated; how are security updates delivered, and can they be delivered remotely; what independent security certifications does the provider hold; is multi-factor authentication supported for platform access; how are permissions structured; and what is the incident notification process if the provider suffers a breach.
Over-the-air update capability deserves specific attention. A device that cannot be updated remotely will not receive security patches in practice, because nobody will bring 300 vehicles into a workshop for a firmware update. Our own approach is set out in the Trust Centre.
It depends on the systems and on whether the supplier has a documented, supported interface. Ask three specific questions rather than the general one:
“Yes, we can integrate with that” is true of almost everything given enough budget. The useful answer is whether it already exists, works, and is supported.
Honestly assessed, more than most business cases assume — and the time is what generates the return. As a rough guide: someone reviewing events and running coaching for a mid-sized fleet is a meaningful part of a role, not a five-minute task. A large fleet with an active video programme typically needs dedicated resource.
The comparison that matters is not “time versus no time”. It is time spent reviewing events against time currently spent reconstructing incidents from scratch, arguing about disputed claims, locating equipment and chasing paperwork. For most operations the new work replaces more work than it creates — but only if someone is actually assigned to do it.
Percentage improvements quoted across this industry vary enormously, because they depend on the starting point, the sector, the vehicle type, how well the programme is run and how the numbers were counted. A fleet with poor baseline performance and no existing safety programme will show dramatic improvements; a well-run fleet will show smaller ones.
Treat any headline figure — from any source, including ours — as an indication of what is achievable rather than a forecast of what you will achieve. Build your own business case on your own numbers, using conservative assumptions. It will be more defensible internally, and it will be right.
Ask us anything — including the questions you think are too basic. We would rather answer them now than after you have bought the wrong thing. Call 0800 020 9339 or email info@visiontelematics.com.
English
Deutsch
Français
Italiano
Español
Nederlands
Português
Polski
العربية
עברית
Afrikaans
Svenska
Dansk
Suomi
Norsk
English (US)
English (AU)
English (NZ)
Français (CA)