Data, Telematics and Measurement

What data a professional contractor should record

Every serious forestry machine should generate commercial data.

Record:

Date
Site
Machine
Operator
Engine hours
Productive hours
Delay hours
Fuel
Production
Maintenance cost
Consumables
Breakdowns

Then calculate:

Fuel per m³
Cost per m³
m³ per productive hour
Repair cost per hour
Availability
Utilisation

For mulching:

ha/hour
L/ha
teeth/ha
$/ha

For chipping:

tonnes/hour
L/tonne
knife cost/tonne

The discipline here is not administrative. Without these numbers, the next machine purchase is a guess, the next contract negotiation is unarmed, and underperforming assets stay hidden inside a fleet average.

Why telematics matters commercially

Modern forestry platforms such as John Deere TimberMatic, Komatsu Smart Forestry, Ponsse Opti systems and Tigercat monitoring systems allow contractors to move toward data-driven costing.

John Deere's H-Series, for example, integrates machine and production information through TimberMatic and mapping systems. Komatsu's Smart Forestry ecosystem allows machine and fleet information to be monitored across one or multiple forest machines.

The commercial value is not the dashboard itself.

The value comes from identifying:

• low-productivity operators
• excessive idle
• fuel anomalies
• long extraction cycles
• unnecessary travel
• downtime patterns

Each of those is a correctable cost. A telematics system that reports them but is never acted upon is an expense, not an investment — which is why the purchase question should be "who in the business will read this weekly, and what will they change", not "does the machine have telematics".

What the data is actually for

Machine data earns its place when it answers a commercial question, and there are only a handful worth asking regularly.

Where does the time go? The gap between scheduled hours, engine hours and productive hours is where utilisation is lost, and it is usually larger than operators or owners estimate. Machines routinely record 15-30% more engine hours than productive hours, and knowing which category the difference falls into — travel, waiting on trucks, breakdown, refuelling, weather — tells you what to fix.

Where is the bottleneck today? Constraints move with haul distance, stem size, weather and truck availability. Data that shows which stage is waiting, and for what, turns a quarterly argument into a weekly adjustment.

Is the cost per unit stable? Fuel per cubic metre, consumable cost per hectare, litres per tonne — these should be steady, so a change is a signal. Tracked per hour instead, the signal disappears into how hard the machine happened to be working.

Is a component heading for failure? Temperature trends, pressure behaviour and recurring fault codes carry warning well before a breakdown, and moving a repair from unplanned to scheduled converts lost production into planned downtime.

The measures worth keeping

Most operations do not need a data programme. They need five numbers, tracked consistently, in units that mean something commercially.

1. Productive hours — separately from engine hours, with delay hours categorised.
2. Production per productive hour — in the unit you are paid in.
3. Mechanical availability — available scheduled hours as a percentage of total scheduled hours.
4. Fuel per unit of output — not per hour.
5. Consumable cost per unit of output — teeth per hectare, chains per cubic metre, hammers per tonne.

Every one of these feeds directly into the cost model. A machine's real cost per cubic metre cannot be calculated without the first two, and the last two are the earliest warning available of a technique, material or maintenance problem.

What telematics does not solve

It does not replace the cost model. Telematics reports utilisation and consumption; it does not tell you whether the machine is earning. That calculation still has to be done, and it needs the residual value and fixed costs that no machine reports.

It does not fix a bottleneck. Data identifies the constraint faster than observation does. Moving it still requires a decision about landings, trucks, shifts or capacity.

It does not substitute for records on a used machine. A data history is useful when it exists and transfers, but service records, ownership history and physical inspection remain the primary evidence.

It is not free. Subscriptions, integration and, most significantly, the time to review the reports are real costs.

Using it without a system

A contractor with two machines does not need a platform. A shared spreadsheet with the five measures above, updated weekly, produces most of the available value, and it has the advantage of being read.

The discipline that matters is consistency of units and definitions — productive hours defined the same way each week, production in the unit you invoice in, consumables costed rather than counted.

See total cost of ownership for where these measures feed into the cost model, and the production data checklist for what to ask a seller or a dealer to produce.

Finding a constraint from data

The most valuable thing a small operation can do with machine data is locate its bottleneck, and it requires less instrumentation than most people assume.

Take a week's records for a full-tree chain:

Two readings fall straight out. The loader is waiting on trucks for nearly half its scheduled time, and the feller buncher is stopping because the landing is full — which is the same problem seen from the other end of the chain. The constraint is haulage, and no machine in the coupe would fix it.

That analysis needed four numbers per machine and a note of what each was waiting on. It did not need a platform, and it is the single highest-value thing an operation can start recording.

Starting without a platform

A two-machine operation does not need a system. It needs five numbers recorded consistently, and consistency matters far more than precision.

Week one: record scheduled hours, productive hours and delay hours with a reason, per machine. Nothing else.

Week four: add production per productive hour in the unit you invoice in. You now have cost per unit for the first time, and it will differ from what you assumed.

Week eight: add fuel and consumables per unit of output rather than per hour. These are the diagnostic measures — they should be stable, so movement means something has changed.

Ongoing: review monthly, not daily. Daily numbers are noise; monthly trends are signal. The review is the point — data nobody looks at is a cost with no return, and that is the most common outcome of telematics projects.

Questions to ask before buying a telematics subscription

Telematics is sold on capability and used on habit. Five questions separate a subscription that will earn from one that will lapse.

1. Does it distinguish productive hours from engine hours? If it only reports engine hours, it cannot answer the question you most need answered.
2. Can delays be coded with a reason, by the operator, without friction? A reason code that takes three screens will not be entered.
3. Can the data be exported? If the figures cannot leave the platform in a usable form, they cannot feed your cost model.
4. Who owns the data, and what happens to it when the machine is sold or the subscription ends? This matters for the machine's service history at resale as well as for you.
5. What will the monthly review actually look like, and who does it? A subscription with no owner inside the business is a subscription that will be cancelled in a year having produced nothing.

What data does for resale

One underrated return: a complete, exportable production and service record makes a used machine easier to sell and easier to defend a price on.

The buyer of a used forestry machine is trying to establish what the machine has actually done — hours, duty, maintenance discipline, faults. A seller who can produce that record removes uncertainty, and uncertainty is what discounts a used machine. Since depreciation is usually the largest hourly cost in the model, anything that narrows the range of resale outcomes is worth more than it appears.

Fuel data is the most underrated measure

Fuel is usually the second or third largest variable cost, it is measured automatically on almost every modern machine, and it is almost never used diagnostically.

Tracked per hour it tells you how hard the machine worked. Tracked per unit of output — litres per cubic metre, per tonne, per hectare — it becomes one of the most sensitive indicators available:

• A step change usually means technique, material or a mechanical problem, and it appears before the mechanical problem becomes a failure.
• Sustained differences between operators on the same machine and material quantify a training opportunity in dollars rather than impressions.
• A gradual drift upward on a processor or chipper frequently means blades or knives are being run past their useful point, because dull tooling raises fuel per tonne before it visibly degrades the product.

None of that requires a platform. It requires recording litres against output rather than against hours, which is a decision about how you write the number down.

Operators and the data

Machine data is also, unavoidably, data about people, and how that is handled determines whether the numbers you get are honest.

Two practical positions worth taking early:

Be explicit about what is recorded and why. Data introduced without explanation is read as surveillance, and the predictable response is delay codes that stop reflecting reality. A delay reason nobody trusts is worse than no delay reason at all, because it produces confident wrong conclusions.

Use it to find system problems, not individual fault. The highest-value findings in this chapter's worked example — a loader waiting on trucks, a feller buncher stopped by a full landing — are system problems that no operator could have fixed. Using data that way, visibly, is what makes operators willing to record the reasons that reveal them.

Benchmark against yourself

There is very little reliable public benchmarking data in Australian forestry, and what circulates informally is usually quoted without the conditions that produced it — species, terrain, haul distance, stem size, operator experience. Comparing your figures against someone else's number is therefore a weak test and frequently a misleading one.

Your own history is a much stronger benchmark. The same machine, the same crew, the same kind of site, measured last quarter, is a comparison where the variables are known. That is what makes consistency of definition matter more than precision: a rough series measured the same way every month tells you more than a precise figure measured differently.