Paving
Intelligent compaction (IC) for asphalt: pass maps, ICMV, temperature
How an IC roller maps every pass, mat temperature, and relative stiffness in real time so you cover 100 percent of the mat to a uniform pattern, find the weak and missed areas while you can still fix them, and keep the data as the as-built record. ICMV is relative stiffness, not density, and cores still accept the work.
Direct answer
Intelligent compaction is rolling asphalt with a vibratory roller that carries GPS, a drum accelerometer, and a temperature sensor, feeding the operator a live color map of pass count, mat temperature, and relative stiffness. It covers 100 percent of the mat, but it supplements core density acceptance rather than replacing it.
Key takeaways
- Intelligent compaction rolls asphalt with a vibratory roller carrying GPS, a drum accelerometer, and an infrared temperature sensor, mapping pass count, mat temperature, and stiffness live.
- ICMV (intelligent compaction measurement value) is relative stiffness, not density; values vary by manufacturer (CMV, Evib, branded index) and do not transfer between machines.
- IC supplements core and gauge density acceptance, it does not replace it; final-pass ICMV correlates poorly with core density, so cores still accept the work.
- IC measures 100 percent of the mat it rolls versus a handful of cores, exposing cold streaks, missed passes, and soft support spots while the mat is still workable.
- Framework is the FHWA IC program and AASHTO standard practice (cited as AASHTO PP81 for soils); confirm current designation and project spec, since standards get renumbered.
Intelligent compaction, and what the roller actually maps
Intelligent compaction is rolling asphalt with a vibratory roller that records what it does and shows it to the operator in real time. The machine carries a GPS receiver, an accelerometer mounted on the drum, and an infrared temperature sensor, and it feeds an onboard display a live color map of three things: how many passes each part of the mat has taken, the surface temperature of that spot, and a relative stiffness reading the industry calls the ICMV. The old method was to roll a pattern, take a few cores or nuclear gauge shots, and trust that the rest of the mat looked like the handful of spots you checked. IC changes that. The operator sees the whole mat in color while there is still time to do something about it.
The gain is coverage. A core tells you about a 4 in or 6 in circle. An IC roller tells you about every square foot of mat it rolled, and it shows where the pattern slipped, where the mat ran cold, and where the support underneath is soft. None of that replaces the density core. Cores and gauge readings still accept the work against the spec. What IC adds is the picture between those test points, which is exactly where most of the trouble hides.
This guide is the technology layer. The companion compaction and rolling guide covers the density target, the temperature window, and the rolling pattern itself, and the paving inspection and quality control guide covers acceptance of the whole pave. IC rides on top of those fundamentals. It sharpens them. It does not stand in for them.
Why intelligent compaction beats spot-checking a few cores
The case for IC is simple arithmetic. A typical day of paving might draw a dozen cores across thousands of square yards. Those cores accept the lot, but they are a tiny sample, and the spots between them are unknown. IC measures the whole mat, every pass, and turns the unknown into a map.
Five things come out of that. You get 100 percent coverage instead of a scatter of points. You get real-time feedback, so the operator corrects the pattern on this mat instead of finding out from a failing core tomorrow. You get a consistent number of passes across the width, which is what density uniformity actually depends on. You find the weak and missed areas while the mat is still hot enough to work. And you get a documented record of how the mat was compacted, which matters when someone questions the pavement later.
The honest framing is that IC does not make the pavement denser by itself. A roller with a screen is still a roller. What it does is take the guesswork out of where you have rolled, how hot the mat was when you rolled it, and where the mat is responding differently than the rest. On a long mainline pull that runs uniform, IC mostly confirms a good pattern. On a job with cold joints, soft subgrade, and a crew chasing the paver, it earns its keep by catching the problems a dozen cores would walk right past.
What is on an IC roller
An IC roller is a vibratory steel-drum roller with a measurement and documentation package added to it. Strip away the marketing and the package is four parts plus the data.
The GPS receiver locates the drum, usually to a few centimeters using a base station or a real-time corrected signal, so every reading is tied to a position on the mat. The accelerometer on the drum measures how the drum bounces against the material, and the onboard computer turns that response into the ICMV. The infrared temperature sensor reads the surface temperature of the mat just ahead of or behind the drum. The onboard display shows the operator a color map built from all of it, updated pass by pass. Some machines add automatic feedback control that adjusts vibration amplitude or frequency to the stiffness it senses, but that is an option, not the definition.
The data file is the fifth part and the one people forget. Everything the machine senses is logged with position and time, and that log is the as-built compaction record. Without it you have a fancier roller. With it you have a document.
| Component | What it does | What it produces |
|---|---|---|
| GPS receiver | Locates the drum on the mat | Position for every reading |
| Drum accelerometer | Senses drum response to the material | ICMV, a relative stiffness value |
| Infrared temperature sensor | Reads mat surface temperature | The temperature map |
| Onboard display | Shows the live color map | Pass count, temperature, ICMV in color |
| Data logger | Records everything with position and time | The IC data file, the as-built record |
What does the pass-count map show?
The pass-count map shows how many times the drum has covered each part of the mat, drawn in color across the full rolling width. It is the most useful single output IC gives you, and the one with the least argument behind it, because pass count is a direct count, not an inference.
Read it and the pattern problems jump out. A stripe that is one color lighter than the lanes beside it took fewer passes, which usually means the operator drifted and left a gap, or overlapped short and missed the edge of the previous lane. A band that ran darker took extra passes, which can over-roll and even break aggregate or push the mat. The goal the map drives you toward is uniform coverage: every part of the mat sees the same number of passes in the same temperature window, because uniform passes are what produce uniform density.
This is where a new operator changes habits fastest. Without the map, you count passes in your head and lose track on a wide mat or a long pull. With the map, the missed strip is right there in front of you, a different color, and you go back and hit it while the mat is still hot. The map turns pass count from a thing you hope you did into a thing you can see you did.
What is ICMV?
ICMV stands for intelligent compaction measurement value, and it is a relative measure of how stiff the material under the drum is, calculated from the drum accelerometer. It is not a density number. Read that twice, because mistaking it for density is the most common and most expensive error in the whole subject.
Here is the mechanism. As the material under the drum gets stiffer, the drum bounces differently, and the accelerometer picks up that change in response. The onboard computer turns the response into a single value. The trouble is that there is no one ICMV. The Federal Highway Administration coined ICMV as a generic, umbrella term precisely because every manufacturer computes its own version with its own math and its own units. Some report a dimensionless compaction meter value, CMV. Others report a vibratory modulus, Evib, in meganewtons per square meter. Others report their own branded index. A number from one system does not transfer to another, and a target set for one machine is meaningless on a different one.
So treat ICMV as a trend, not an absolute. Rising ICMV over successive passes tells you the material is stiffening up, which is what you want to see. A low ICMV area sitting next to a high one tells you something is different there worth investigating. What ICMV does not tell you, on its own, is the in-place density in percent of maximum. For that you take cores. The value of ICMV is in the relative picture across the mat and the trend across passes, and that value is real, as long as you never let anyone read it as density.
Temperature mapping and the rolling window
The infrared sensor maps mat surface temperature as the roller works, and that map is arguably more directly useful than the stiffness map, because temperature drives compaction in a way that is well understood. Asphalt only takes compaction inside a temperature window. Roll it too hot and the mat shoves and checks. Let it cool past the low end and the mix stiffens up and stops taking density no matter how many passes you make. The compaction and rolling guide covers that window and the target by mix; here the point is that IC shows you the temperature everywhere, not just where someone happened to point a handheld gun.
The big find is thermal segregation, the cold spots and cold streaks that come off the paver or out of the truck and that a handful of spot readings will miss. A cold streak down the mat is a strip that will never reach density because it left the window before the roller got there, and on the temperature map it shows as a distinct color you can chase back to its source: a paver that sat, an end-of-load drop, a wind-exposed edge, a cold truck.
Pairing the temperature map with the pass-count map is where the real discipline lives. Passes only count if they land in the window. A spot that took plenty of passes but took them after the mat went cold is still a weak spot, and only the two maps together tell you that.
Reading the color maps
All three outputs reach the operator the same way, as color maps on the cab display drawn over the geometry of the mat. Pass count, temperature, and ICMV each get their own view, and the operator and the QC tech read them the way a foreman reads a set of plans: not for the pretty picture, but for the spot that is the wrong color.
The color is relative to a scale, and that is the thing to understand before you trust it. Pass count is plain, the colors step up with each pass, and there is nothing to interpret. Temperature is anchored to real degrees, so the color means a number. ICMV color is relative to a range the system or the operator sets, so a green spot is greener than a red spot, but green is not a density and not an acceptance, it is just stiffer than red on this machine today.
On the job the maps get used two ways. Live, the operator watches the pass-count and temperature maps and steers the pattern. After, the QC tech and the inspector pull up the ICMV and temperature maps for the lot and look for the uniformity, the cold areas, and the soft spots that deserve a core. The map is a tool for finding where to look, not a substitute for what you find when you look there.
The operator, the map, and real-time correction
The defining feature of IC is that the information reaches the operator while the mat is still workable. This is feedback, and the human in the seat is the one acting on it. Even on machines with automatic amplitude control, the operator is still steering, still deciding to go back over a light strip, still calling whether a cold area is worth chasing. IC is not autopilot.
What changes is what the operator is working from. Without the map, a roller operator works from a pattern set in a pre-pave meeting, a pass count held in memory, and a feel for the mat. With the map, the operator works from a pattern and the live picture of how well it is being executed, which catches the drift and the missed overlap that memory does not. The good operators take to it fast, because it confirms what their hands already told them and shows them the spots their hands could not feel.
The failure mode is an operator who watches the screen instead of the work, or who chases every color flicker into over-rolling. The map informs the pattern. It does not replace the pattern, and it does not replace the operator's judgment about what the mat is doing.
100 percent coverage and the shift in how QC works
Conventional acceptance is a sampling problem. You cannot core everywhere, so you core a few spots, run the statistics, and accept the lot on the sample. That works, and it is still how the work gets accepted, but it leaves the space between cores unmeasured. IC measures that space. Every square foot the roller covers gets a pass count, a temperature, and an ICMV.
The practical shift is from accept-the-sample to find-the-outlier. With a coverage map, you are not limited to asking whether the random cores passed. You can ask where the mat is different, and then go core or gauge that spot on purpose. A soft area on the ICMV map, a cold streak on the temperature map, a light strip on the pass map: those are the places a targeted test earns its keep, and they are places a random core would find only by luck.
Several owners now use IC exactly this way, as a quality control map that directs where the acceptance tests go, rather than as the acceptance test itself. That is the right altitude for it today. The coverage finds the suspects. The cores convict them.
Does ICMV measure density?
No. ICMV is a relative stiffness response from the drum, and it is not a density measurement. It correlates with density only roughly and only under the right conditions, which is the single fact that has kept IC from simply replacing cores, and the fact most likely to get glossed over by someone selling the technology.
The research is consistent and worth knowing. Pass-by-pass ICMV during breakdown rolling can track density reasonably well, which is why it is useful for quality control as the mat builds. The ICMV on the final pass, the one you would want for acceptance, does not correlate well with core density. The reasons are physical: the drum senses down through the new lift into whatever is under it, so a stiff base reads as high ICMV regardless of the mat on top, and once the mat cools below a threshold the drum response changes again. On a thin lift the underlying support dominates the reading entirely.
So the rule is firm. Take cores for acceptance. Use ICMV to see uniformity and trend and to find the spots that deserve a core. Anyone who points at a green ICMV map and calls the density good has skipped the one step the data cannot do for them. ICMV is relative, it varies by system, and it is not a direct density number, every time, on every job.
Building the ICMV-to-core correlation
Because ICMV is relative, it only means something once you tie it to a number you trust, and that is done with a correlation built on a test section. The idea is to roll a representative strip, record the ICMV, take cores or gauge readings at marked spots, and look at how the two track. That relationship, for this machine, this mix, this support, and this day, is the calibration that lets you read the ICMV map with any confidence.
The catch is that the correlation is weak more often than anyone wants. Published thresholds for an acceptable correlation coefficient sit around 0.5 to 0.7, and plenty of test sections do not reach it, especially on thin lifts and over variable support. When the correlation does not hold, the answer is not to force it. The answer is to use ICMV for what it still does well, which is showing relative uniformity and trend across the mat, and to keep accepting on cores.
Build the correlation on the same support and the same mix as the production work, not on a convenient pad with different conditions underneath. A correlation built over a stiff base does not transfer to a soft one, because the ICMV will read differently for reasons that have nothing to do with the mat. Rebuild it when the conditions change. Treat the number it gives you as approximate, and verify the values against the project specification and the AHJ before you lean on them for anything that matters.
What does an IC specification require?
An IC specification typically asks for some mix of three things: a target number of passes across the mat, a coverage percentage, and in the more advanced specs an ICMV target tied to a correlation. The exact requirements vary widely by agency and by spec generation, so read the one on your project rather than carrying a number in your head from the last job.
The reference framework is the Federal Highway Administration IC program and the AASHTO standard practice for IC, commonly cited as AASHTO PP81 for embankment and subgrade soils, along with the FHWA generic IC special provisions for asphalt. Confirm the current designation, because AASHTO has been renumbering its provisional standards and the spec you are held to may reference a newer number or a state-specific version. Many state DOT specs build on that framework and add their own pass-count and coverage targets and their own data-submission rules.
Where acceptance is concerned, most specs today still accept density on cores or gauge and use IC for coverage and uniformity, with pass count and temperature in the window as the enforceable IC requirements and ICMV as a quality control measure. A spec that ties acceptance to an ICMV target requires a validated correlation behind it, and that is exactly where you hedge hardest: the target, the pass count, and the ICMV values all belong to the specific system, the specific spec, and the AHJ, and acceptance built on ICMV is still evolving. Read the spec, run the correlation it calls for, and confirm the targets with the agency before you sign to them.
Where IC started: soils and base
IC did not start on asphalt. It started on soils and subgrade, where rollers were fitted to map the stiffness of embankment and base as a proof of compaction, and the same idea carried up into the mat. On earthwork the value is even cleaner, because there is no temperature window to fight and the stiffness reading is not confused by a hot, changing material on top of a cold base.
The earthwork application is proof-mapping. You roll the layer, the ICMV map shows you where the soil is soft, and you go work those areas before you build on them. A soft spot in the subgrade is a soft spot under everything you put on top of it, so finding it with a map beats finding it later as a rut. The soil IC specs reference soil test methods such as the standard Proctor relationships rather than asphalt cores, and the correlation is built against spot tests like a nuclear gauge or a plate load.
For a paving crew the relevance is the support under the mat. If the base was IC proof-mapped and the soft areas were fixed, the mat above has a better chance of uniform density. If it was not, those soft areas show up again in the asphalt ICMV map, and now they are harder to fix. The base guide covers earthwork compaction in its own right; the link here is that the mat is only as uniform as what it sits on.
Using the map to find and fix weak spots
The highest-value thing IC does day to day is point at the weak spots, and a paving crew that uses it for nothing else is still ahead. A low ICMV area on the map is a flag: the material there is responding softer than the rest, and the reason is worth running down before the mat cools.
Work it like a diagnosis. A low ICMV that lines up with a cold area on the temperature map is a temperature problem, and the fix is to roll it sooner next time, not to beat it cold. A low ICMV that lines up with a light area on the pass map just needs the passes it never got. A low ICMV that sits over good passes and good temperature is the one to take seriously, because that usually means soft support under the mat, and no number of extra passes on top will fix a problem that lives underneath. Mark it, core it, and find out what is down there.
The mistake is to chase every soft area with extra passes regardless of cause. Over-rolling a mat that is the right temperature can break aggregate and decompact it, and pounding a spot that is soft because the base is soft just wastes the window on the rest of the mat. The map tells you where to look. Your judgment, and a core, tells you what to do about it.
The IC workflow on a paving day
IC adds steps to the morning, and skipping them is how a crew ends up with a machine that records nothing useful. The work goes in order: set up the machine, establish the GPS reference, set the plan, roll to the map, and review the data with QC.
Setup means the machine is calibrated, the temperature sensor is reading true, and the ICMV scale is set for the day. The GPS reference means a base station or a corrected signal that puts the drum where it actually is, because a map drawn on a drifting position is worse than no map. The plan means the rolling pattern and pass target from the pre-pave meeting, ideally with a test section to set the pattern and build any ICMV-to-core correlation the spec wants. Then the operator rolls to the pass-count and temperature maps, holding the pattern and chasing the cold and light areas while the mat is workable.
The step that gets dropped is the review. At the end of the lot the QC tech and the inspector pull the data, look at coverage and uniformity, and decide where the targeted tests go. If nobody opens the file, the most useful part of the system went straight to the trash, and you paid for a roller with a screen.
The IC data file as the as-built record
The IC data file is the part with the longest reach, because it outlives the paving day. Every pass, every position, every temperature, and every ICMV is logged, and that log is an as-built record of how the mat was compacted. The standard way to view and analyze it is purpose-built software, with Veta being the analysis tool the FHWA developed so agencies and contractors read the data the same way; the AASHTO practice also defines a tagged file format so data moves between systems.
That record is worth real money in a dispute. When a pavement is questioned a year out, the contractor who can show a coverage map proving the mat got its passes in the window stands on solid ground, and the one with a dozen cores and a memory does not. The record also feeds the owner the proof they increasingly ask for, and it tells the contractor's own QC where the process drifted across a season.
This is where the field record-keeping pays off, and where a tool like FieldOS fits: capturing the IC outputs, the test-section correlation, the cores that went with the lot, and who reviewed the data, all tied to the lot and the date, so the as-built is one organized record instead of a data file on a USB stick that nobody can find when the question comes. The map is only proof if you can produce it on demand.
The limitations you have to respect
IC has real limits, and the crews that get the most out of it are the ones honest about them. The headline limit is the one already covered: ICMV is relative, it is not density, and the final-pass reading correlates poorly with cores. The rest follow from how the measurement works.
The drum senses through the lift into the support beneath it, so the ICMV reflects the whole stack, not the mat alone. On a thin lift the support dominates and the mat barely registers. A stiff base reads high even over a marginal mat, and a soft base reads low even over a well-compacted one. Mat temperature changes the drum response, so the same density reads differently hot and cool. Moisture and material changes move the number on soils. And the value carries no units you can compare across machines.
The discipline that comes out of all this is one sentence: do not over-trust a single ICMV number. A single point means little. The trend across passes means something. The relative pattern across the mat means something. A spot that disagrees sharply with its surroundings means go look. Anyone treating one ICMV value as a verdict has misunderstood the tool, and the spec and the AHJ are the authority on what the value is allowed to decide.
Training the operator and the QC to act on the maps
The technology is only as good as the people reading it, and an IC roller run by an operator who never learned to read the screen is an expensive conventional roller. Training is the part that separates a job that gets value from IC from a job that just has the equipment on it.
The operator has to learn to hold the pattern to the pass-count map without staring at the screen instead of the mat, to read the temperature map and roll the cold areas first, and to resist over-rolling a spot that is soft for a reason the drum cannot fix. The QC tech and the inspector have to learn to read the coverage and uniformity, to target tests at the outliers the map reveals, and to understand exactly what ICMV is and is not, so nobody in the chain accepts a mat on a color.
This is also where the pre-pave meeting earns its place. The pattern, the pass target, the test section, and the plan for what happens when the map shows a problem all get settled before the paver moves, not argued over on a cooling mat. The tech does not train the crew on its own.
Factory IC versus retrofit kits
There are two ways onto an IC roller. The machine comes from the factory with the package integrated, or a retrofit kit is added to a conventional roller. Both put GPS, temperature, and a display on the machine and log the data, and both can satisfy a spec, so the choice is about fit and money, not capability.
Factory systems, the BOMAG, Hamm, Sakai, Caterpillar, Dynapac, and Volvo offerings among them, integrate the accelerometer with the machine's own vibration system and tend to support automatic feedback control and the manufacturer's own stiffness metric cleanly. Retrofit kits from suppliers such as the GPS and machine-control vendors add mapping and documentation to a roller you already own, which is the cheaper path onto IC and often the practical one for a fleet that is not ready to replace machines. What a retrofit may not give you is the integrated stiffness feedback that comes from a system designed around the machine.
Match the system to what the spec actually requires. If the spec wants pass-count and temperature mapping with data submission, a retrofit kit covers it. If it wants a validated ICMV correlation and feedback control, the integrated factory system is the surer route. Confirm that whatever is on the machine produces the data format the agency wants before the job, not after.
Where the agency specs are heading
More DOTs are specifying or incentivizing IC every cycle, and the trend has been one direction for years: from optional, to incentive-paid pilot work, to required mapping with data submission on more and more projects. Several states have moved to make IC standard on mainline paving, and the federal program has pushed the framework and the analysis tools to make that consistent across agencies.
What has matured fastest is the coverage and temperature side. Pass-count mapping and temperature mapping are well proven, and specs that require them are on solid ground. What is still settling is acceptance on ICMV, which is held back by the correlation problem and is best read as evolving rather than arrived. Read the trend as the mapping becoming routine while the stiffness-based acceptance stays cautious.
For a contractor the takeaway is to expect IC requirements to keep spreading and to get the workflow and the data handling sorted before a project mandates it, because learning the system on a job with money tied to the data is the hard way to learn it. Confirm each agency's current spec and version; this field changes spec to spec and year to year.
What IC is worth
The return on IC is not a denser mat for free, and selling it that way sets the wrong expectation. The return is uniformity, fewer weak areas that fail early, less rework caught while the mat is still hot instead of after it cools, and a documented record that defends the work and satisfies the owner. Those are real, and on the right job they cover the cost of the equipment and the workflow.
Where it pays best is the job with variable conditions: changing support, lots of joints, long hauls with temperature loss, the work where a dozen cores are most likely to miss the problem. Where it pays least is short, simple, uniform work that a good crew already controls by hand. The equipment cost, the training time, and the data handling are real line items, so the honest answer on ROI is that it depends on the job and the agency incentive, and it is strongest where the risk of a non-uniform mat is highest.
Count the documentation as part of the value, not a side effect. On work where the owner asks for proof, or where disputes are likely, the as-built record is worth carrying on its own.
IC supports the fundamentals, it does not replace them
The thing to keep straight is that IC sits on top of conventional rolling, it does not replace it. The density target, the temperature window, the three rolling phases, the roller types, and the pattern set on a test section are all still the work, and they are covered in the compaction and rolling guide. IC measures how well you executed those fundamentals. It does not do them for you, and it does not change what good looks like.
Acceptance is the same story. Cores and gauge readings still accept the mat against the spec, exactly as the inspection and quality control guide lays out. IC does not retire the core. It tells you where to take a smart one and proves the coverage between them. Use both. The map without the cores is unverified, and the cores without the map are blind to everything between them.
The crews that get it wrong treat IC as either a gimmick to ignore or a magic box that replaces testing. It is neither. It is a measurement layer that makes a good rolling operation visible and a sloppy one obvious, and it is worth exactly as much as the fundamentals it sits on.
Common mistakes
- Treating ICMV as a density number instead of a relative stiffness reading that varies by system.
- Skipping the test section and never building an ICMV-to-core correlation, then trusting the ICMV map anyway.
- Ignoring the temperature map and rolling cold streaks that will never reach density no matter how many passes.
- Putting an operator on the machine with no training to read the maps and act on them.
- Over-trusting a single ICMV value, or chasing every color flicker into over-rolling.
- Pulling the data file once and never reviewing it, so the coverage map and the as-built record go to waste.
- Adding extra passes to a soft spot caused by soft support, when the problem lives under the mat, not in it.
What to document
The IC outputs are only proof if you keep them organized and tied to the lot. Capture the maps, the correlation, and the conventional tests together, so the as-built tells the whole story when someone asks for it a year later.
Record the maps and the metrics that go with each one, the test-section correlation and the conditions it was built on, the cores or gauge readings that accepted the lot, the spec targets you held to, and who reviewed the data. The map proves coverage. The cores prove density. The note proves someone looked.
| Map or metric | What it shows | Note to keep with it |
|---|---|---|
| Pass-count map | Coverage and pattern uniformity | Target pass count and any light or missed areas reworked |
| Temperature map | Mat temperature during rolling | Window held, cold streaks and their cause |
| ICMV map | Relative stiffness across the mat | Relative only, not density; the scale used |
| Test-section correlation | ICMV against cores for this setup | Coefficient reached, mix and support conditions |
| Cores or gauge readings | In-place density for acceptance | Locations, results, lot accepted against the spec |
| Data file and reviewer | The as-built compaction record | Software used, who reviewed, date and lot |
Field checklist
Want this checklist to run itself on every job — with photo proof and a signed record crews can hand the customer? That's FieldOS.
Standards and references
The framework comes from the Federal Highway Administration IC program and the AASHTO standard practice for intelligent compaction, commonly cited as AASHTO PP81 for embankment and subgrade soils, together with the FHWA generic IC special provisions for asphalt and for soils. The AASHTO practice also defines a tagged data file format so IC data moves between systems, and the FHWA-developed Veta software is the common tool for viewing and analyzing it. Confirm the current designation before you cite it, because AASHTO has been renumbering its provisional standards and a project may reference a newer number or a state-specific version.
The numbers that matter are not in those documents in a form you can carry job to job. The ICMV is a relative value that varies by system and manufacturer, whether it is reported as CMV, Evib, or a branded index, so the ICMV target, the pass count, and any coverage percentage belong to the specific system, the specific spec, and the AHJ. Hedge all of them to the project, run the correlation the spec calls for, and verify the targets with the agency.
Density acceptance stays with the conventional standards: cores and the in-place density methods your DOT spec and the Asphalt Institute reference, accepted against the mix design and the project specification. The discipline holds across all of it: cover 100 percent of the mat and map the temperature, correlate ICMV to cores and still take the density, and use the maps to find and fix the weak spots. ICMV is relative stiffness, not density, every time.
Units and terms
IC carries its own vocabulary, and the same idea reads differently across a spec, a manufacturer sheet, and a software report. The terms below are the ones that cause the most confusion on the job.
ICMV is the umbrella term; the specific value on a given machine may be a dimensionless CMV, a vibratory modulus Evib in meganewtons per square meter, or a manufacturer's own index, and none of them convert to another. Pass count is a plain count. Temperature is read in degrees Fahrenheit or Celsius depending on the system. Correlation is reported as a coefficient, where a value closer to 1 means the ICMV and the cores track more closely.
- Intelligent compaction (IC)
- Rolling with a vibratory roller equipped with GPS, a drum accelerometer, a temperature sensor, a real-time display, and data logging that maps the work
- ICMV
- Intelligent compaction measurement value, a relative stiffness response from the drum accelerometer that varies by system and is not a density number
- Pass-count map
- A color map of how many times the drum covered each part of the mat, used to hold uniform coverage
- Temperature mapping
- A color map of mat surface temperature from the infrared sensor, used to keep rolling in the window and find cold spots
- Correlation
- The relationship between ICMV and core density built on a test section, reported as a coefficient, that lets the ICMV map be read with confidence
- GPS
- Position from a base station or corrected signal that ties every IC reading to a location on the mat
FAQ
What is intelligent compaction?
Intelligent compaction is rolling asphalt or soil with a vibratory roller fitted with GPS, a drum accelerometer, a temperature sensor, and an onboard display. It maps pass count, mat temperature, and relative stiffness in real time across 100 percent of the mat, so the operator and QC see coverage instead of guessing between cores.
What is ICMV?
ICMV stands for intelligent compaction measurement value, a relative measure of material stiffness calculated from the drum accelerometer. The FHWA coined it as an umbrella term because each manufacturer computes its own version, such as CMV or Evib, with its own units. It is a trend and a relative pattern, not a density number, and does not transfer between systems.
What does intelligent compaction measure?
An IC roller measures three things tied to GPS position: how many passes each part of the mat took, the mat surface temperature from an infrared sensor, and a relative stiffness response from the drum accelerometer, the ICMV. It maps all three in color in real time and logs them as an as-built record of how the mat was compacted.
Does intelligent compaction replace density testing?
No. IC supplements core and gauge density acceptance, it does not replace it. ICMV is a relative stiffness reading that correlates poorly with core density on the final pass, so cores still accept the work. IC adds 100 percent coverage and temperature mapping that show where to take a smart core and prove the coverage between them.
Is ICMV the same as in-place density?
No. ICMV is relative stiffness from the drum, not density in percent of maximum. It correlates with density only roughly, and the final-pass reading correlates poorly because the drum senses through the lift into the support beneath and reacts to temperature. Build an ICMV-to-core correlation on a test section and still take cores for acceptance.
How much of the mat does intelligent compaction cover?
An IC roller measures 100 percent of the mat it rolls, every square foot, versus a conventional job that accepts a lot on a handful of cores. That coverage finds the cold streaks, missed passes, and soft areas a random core would miss. The cores still accept the work; the coverage map tells you where to point them.
What does an IC roller do that a conventional roller does not?
A conventional roller compacts; an IC roller compacts and maps. It adds GPS, a drum accelerometer, a temperature sensor, a live color display, and data logging, so the operator sees pass count and temperature across the whole mat in real time and corrects the pattern while the mat is still workable, and the data becomes an as-built record.
Do I still need cores if I have intelligent compaction?
Yes. Cores and gauge readings remain the density acceptance, because ICMV is relative stiffness and does not reliably equal density. Use IC to find the weak, cold, and missed areas and to target where the cores go, then accept the lot on the cores against the project spec and the AHJ. Use both together, not one instead of the other.
What IC spec applies to asphalt paving?
The framework is the FHWA IC program and the AASHTO standard practice, commonly cited as AASHTO PP81, plus FHWA generic IC special provisions, with state DOT specs adding their own pass-count and coverage targets. Confirm the current designation and the project version, because the targets and acceptance rules vary by agency and the standards get renumbered.
Why does an ICMV reading show low in one spot?
A low ICMV means the material there responds softer than the rest. Check the other maps: low ICMV over a cold spot is a temperature problem, over a light area is missed passes, and over good passes and good temperature usually means soft support under the mat. Mark it, core it, and fix the cause, not just the symptom.