Two fleets instrument their vehicles and compare the resulting wear. One finds that the machines with the highest recorded hours consume the most edges, and concludes that hours are the driver; the other finds almost no relationship between hours and wear, and concludes that the data is useless. Both are reading the same kind of data, and neither has yet looked at what the vehicles were doing with the blade down.
This is a field Q&A on what vehicle data can contribute to blade management: which streams contain information about wear, which do not, how the data is combined with physical inspection, how to set up a study that produces a usable answer, and what to expect from it in the first season.
What fleet data can and cannot tell you
Vehicle data describes the machine’s activity, and blade wear is caused by the machine’s interaction with a surface. That gap is the whole of the limitation, and it is why data alone has never replaced a measurement.
What the data can do is narrow the interval over which a measurement has to be taken. If a fleet knows that a particular machine has accumulated twice the blade-down moving time of another, it knows that its edge should be inspected sooner, and it can schedule that inspection rather than waiting for a calendar date. What the data cannot do is tell the fleet how much material is left, whether the seat is fretting, or whether the route has changed from one that abrades to one that does not.
Those three questions remain measurement questions, which is why a data programme sits alongside an inspection routine rather than replacing it.
Why the data feels more conclusive than it is
Telematics output has a quality that makes it easy to over-read. It is continuous, it has timestamps, and it produces charts, all of which give the impression that the underlying question has been answered. Wear data has none of those properties: it consists of a handful of measurements per season, taken at marked positions by a person with a rule.
The practical consequence is that a fleet can end up optimising the data rather than the blades. A report showing which machines accumulate the most contact time is useful; the same report being treated as a wear prediction is not, and it leads to replacement decisions taken on hours rather than on condition. The two should agree broadly, and where they disagree the measurement wins.
Telematics plow blade management: what the data streams contain
Separate blade-work streams from shift streams.
The useful distinction is between streams that record work the blade has done and streams that record the vehicle’s general activity. Fuel, engine hours and total distance describe the shift, and they correlate with blade wear only indirectly.
Position, blade position, speed and time in contact describe the work the blade has done, and those are the streams a study should be built from. Treating the two groups as equivalent is the most common analytical error in this area.
Distance, speed and surface correlation
Within the second group, the correlations behave differently and it is worth being specific about which is useful.
Total distance travelled correlates poorly with blade wear, because a machine can travel a long way with the blade raised. Speed correlates with wear in the sense that it changes contact time at each point, but its effect is nonlinear and depends on the surface, so a speed figure without a surface context supports only a weak conclusion. Blade-down moving time is the strongest of the readily available streams, because it is the closest proxy for the amount of contact the edge has had.
The missing variable across all three is abrasiveness. Two routes worked at the same speed for the same time produce different wear if one is gravel and the other is clean pavement, and no vehicle sensor distinguishes them. That is why the surface has to be recorded externally — from the route definition, from the operator, or from the inspection — and joined to the data rather than derived from it.
What sensors cannot measure
The limits of a vehicle data programme come from the things it cannot observe, and naming them prevents the programme from being asked to do more than it can.
- The remaining dimension of the edge or the condition of the insert array.
- The condition of the seat, the pocket or the mounting line.
- The abrasiveness of the surface the blade has been working.
- Whether an impact occurred that changed the machine’s geometry.
- Whether the operator adjusted the blade, the shoes or the down pressure during the season.
The fifth item is worth recording deliberately, because it is an input the fleet controls and the data will not reveal. A mid-season setting change alters the wear rate of every machine in a group, and without a note the change appears in the data as an unexplained step.
Combining telematics with physical inspection
The two sources answer different questions, and the combination is what produces a maintenance decision.
Telematics says how much blade-down work a machine has done since the last inspection, which sets when the next measurement is due. The measurement says how much material has gone over that work, which produces a wear rate. The rate, applied to the minimum dimension, gives the remaining life. The surface record explains why one machine’s rate differs from another’s.
Run in that order, the data programme reduces the number of measurements a fleet has to take without reducing the information it gets, because the measurements are taken at points where they will show something. The measurement routine that the programme drives is described on the blade wear measurement page.
Pilot design for a blade study
A first study should be small, bounded and designed to produce a correlation rather than a prediction.
- Choose a small number of machines that work different route types and have comparable blade specifications.
- Measure every edge at the start, at marked positions, and record the measurement with the machine.
- Record the route type each machine works, including the surface description.
- Run a fixed period — a month or a fixed number of events, not a season — and then measure again.
- Compare the wear measured against the blade-down moving time recorded for the same period.
The fourth step is what makes the study analysable. A season-long pilot accumulates too many changes to attribute anything, while a month produces a comparison in which the settings, the routes and the operators have all stayed broadly constant. Where the correlation is visible in one month, the following season can extend it.
Privacy and labour considerations
Vehicle data frequently describes individual working patterns, which is why a blade study can be read as performance monitoring whether it is intended that way or not.
- Aggregate the data before it reaches a maintenance decision, so the output is about machines and routes rather than about individuals.
- State in advance what the data is being collected for, and what it will not be used for.
- Involve operators in the study design, since they supply the surface and setting information the data cannot.
- Report the study’s findings back to the crews, as the observations they contributed are part of the result.
The fourth item is what keeps the programme running. A data study that draws on operator knowledge and reports nothing back becomes something the crews work around, and the quality of the surface and setting information degrades with it.
Realistic expectations
A first-season study produces a correlation between blade-down work and wear, expressed per route type. It does not produce a wear prediction, a replacement interval, or a per-machine forecast, and expecting those in the first year is what leads fleets to conclude that the data does not work.
What it does produce is worth having. A fleet that knows which routes consume edges fastest can allocate its stock and its replacement schedule accordingly, and it can identify the machines whose rate is out of line with their route group — which is usually the first sign of a fitment problem rather than a specification one.
What to do before instrumenting the fleet
Three things are worth establishing before a study is designed, and each of them is a prerequisite rather than an optional extra.
- A measurement routine that produces comparable readings at marked positions.
- A route description that records surface type, since the data cannot supply it.
- A record of setting changes made during the study period, including shoe heights and down pressure.
Without the first, the study has no dependable output to correlate against. Without the second and third, any relationship the data shows will be unexplainable, which is the same as unusable.
Where the data pays for itself
Three benefits justify the effort, and none of them requires the data to predict wear. Targeted inspection, because measurements are taken where they will show something. Route-level allocation, because the routes that consume edges fastest can be identified and matched with the right specification. And the detection of outliers, because a machine whose rate is out of line with its route group is a maintenance finding rather than a specification one.
Those three are the reasons to instrument a fleet, and they are all available in the first season. The mid-season audit is where the findings are acted on, and how that is run is described on the mid-season audit page. Fleet technology and operations practice is published by the Snow and Ice Management Association and the National Association of Counties, network practice by the American Association of State Highway and Transportation Officials, equipment standards by SAE International, and road weather context by the Federal Highway Administration.
SENTHAI manufactures plow edges, wear plates and carbide inserts in Rayong, Thailand, and produces edges to the pattern each machine requires, which is what makes a measurement taken on one unit comparable with the same position on another. Studies that need the edge specification and pattern recorded against each machine can be discussed through the contact page.
FAQ
Can telematics predict when a plow blade will need replacing?
Not on its own. Telematics measures what the vehicle did, and wear depends on what the blade met, which no vehicle sensor sees. Used with periodic physical measurements it sharpens the interval considerably, but the measurement is still what establishes how much material is left.
Which data stream correlates most closely with blade wear?
Time in contact with the surface while moving, rather than total distance or engine hours. A machine that spends its shift with the blade raised contributes hours without wear, so the figure that matters has to be the one associated with the blade being down and the vehicle moving.
What can vehicle sensors not measure about a blade?
The condition of the edge, the condition of the seat, and the abrasiveness of the surface being worked. Sensors record position, speed, load and time; none of them can distinguish abrasive gravel from clean pavement, which is the single largest variable in wear.
How should a telematics blade study be set up?
By instrumenting a small number of machines on known routes, measuring their edges at the start and end of a fixed period, and comparing the wear measured against the data recorded. The study produces a correlation, not a prediction, in the first season.
Does a telematics study raise privacy issues?
It can, because vehicle data frequently describes individual working patterns and can be read as performance monitoring. Two practices avoid most of the difficulty: aggregating data before it is used for maintenance decisions, and stating in advance what the data is for and what it will not be used for.
Send the routes, the machines and the edge specification each one runs. The SENTHAI engineering desk will confirm the measurements that pair with the vehicle data so a first-season study produces something usable.