Asphalt Intelligence Engine
Asphalt Technologies
Materials laboratory & pavement forensics · South Carolina, North Carolina, Georgia

The laboratory report that holds up
when somebody asks how you know.

Mixture design, binder verification, recycled materials, production support and pavement forensics — run on a documented engine, delivered with the arithmetic, the provenance of every number, and a clear statement of what the evidence does not support.

Open the app Talk to the laboratory
For a state DOT

Referee work that ends the argument

Independent re-testing with the custody trail attached, and an honest account of how much of the disagreement the method's own precision explains.

For a producer

A design you can defend at the plant

Volumetrics, balanced mix design, RAP and rubber — each with the bound, the assumption behind it and the test that would settle which applies.

For an owner or counsel

Forensics built to survive cross-examination

Ranked causes with what would refute each, the tests that separate them, and the treatments that would bury the mechanism instead of fixing it.

19
Distresses
76
Mechanisms
153
Candidate causes
224
Test specifications
176
Remedies
01

What we do, and what you get back

Seventeen laboratory processes, from the specimen at the door to the witness stand. Every one of them returns a signed-ready report with the inputs as supplied, the arithmetic, the provenance class of every value, and the condition under which it declined to conclude.

Acceptance & disputes

When a lot is contested

  • Independent referee testing with custody per specimen
  • Quantities, yield and lift thickness with the arithmetic shown
  • Causal chain to something a person controlled — never to a party
  • FMEA ranking what to change so the lot does not recur
You get a document your own engineer signs. We issue no pay factor, no acceptance and no rejection: the determination stays with the agency.
Design & materials

Before it is placed

  • Superpave and Marshall design, and JMF verification
  • Balanced mix design — with the width of the region, not just a number
  • RAP and RAS bounded between full blending and none
  • Crumb rubber and polymer qualification, with the confounds named
  • PG grading, MSCR, and the grade the site requires
You get the design and the sentence most laboratories leave out: what it would take for this to be wrong.
Forensics & research

After something has failed

  • Assessment from photographs, cores and report pages
  • Coring plans built against hypotheses, not on a grid
  • Rehabilitation selection, with what each option does not fix
  • Contract research with the specimen count derived before it is priced
  • Expert testimony support and cross-examination preparation
You get a position you can hold, or a documented refusal that is worth more than a guess.
02

Three things a laboratory usually will not tell you

All three are real outputs from this platform, on real cases, quoted rather than described.

The one that saves money
The cheapest repair was the one that hides the failure.

Life-cycle cost ranked a thin overlay at $48,700 a year against $51,200 for mill and inlay. The same overlay is what the forensic engine had already marked as leaving the mechanism running. Cost does not know the mechanism is still alive.

TREATMENTS THAT WOULD HIDE THIS, NOT FIX IT Rut filling with a levelling course · Microsurfacing Structural overlay · Subgrade undercut
The one that protects the file
It refused to compute from impossible numbers.

Handed a design sheet with the bulk and maximum specific gravities transposed, it returned no volumetrics at all. A laboratory that computes air voids from an impossible pair has put a number in a file that somebody will later have to explain.

Gmb > Gmm is physically impossible; re-check Rice or bulk SG
The one that keeps a design honest
A balanced region one point wide is not a margin.

A four-point sweep produced exactly one qualifying mixture. Rather than call that an optimum, the engine reported no region: in production the binder content moves, and a design sitting on a single measured point moves with it.

NO BALANCED REGION. One trial qualifies, at 5.40 %. A single trial is an edge, not a region.
The one that finds the real culprit
The distress is where the failure shows, not where it starts.

Asked about a pothole, it declined to name a root cause inside the pothole record and pointed at the records that actually hold it — structural capacity, moisture damage, drainage — because a pothole is a symptom of all three.

WHERE THE ROOT BEHIND THIS IS HELD weak_or_saturated_base_support -> moisture_damage / drainage deficiency
03

Why it holds up

Four refusals, enforced in the software rather than in a policy anyone can be talked out of. They are the reason a report from this laboratory can be handed to an agency reviewer or taken into a deposition.

Never

Call a symptom a diagnosis

Symptom, mechanism, immediate cause and root cause are separate types in the code, and the chain has to end at something a person can change.

Never

Quote a specification from memory

Every standard is cited by designation and carries its verification state. What a clause requires is not stated without the document in hand.

Never

Name a cause the evidence cannot separate

When the leading explanations sit too close, the answer is an unresolved uncertainty and the test programme that would settle it.

Never

Recommend a repair that leaves the mechanism running

Every remedy declares what it does not fix, so the treatments that would bury the problem are printed as exactly that.

04

Try it before you call us

The laboratory bench is open. Seventeen processes, no account, nothing stored on this page — upload a photograph of a distress or a page of a laboratory report and watch it work, including the moments where it declines to answer.

Forensic diagnosis · P-12 Photograph · core · report page

Four stages run in order: the image is read, the field questionnaire is answered, the deterministic engine ranks the causes, and only then is prose written over the conclusion it reached. The ranking is never the model's.

Open the forensic bench
The seventeen processesLive

Mixture design and balanced mix design. RAP, RAS and rubber. Binder grade from the site. Acceptance quantities and yield. Decision matrices with their robustness tested. Experiment design with the specimen count derived. Root cause and FMEA. And a panel of twenty-seven declared specialists over any deliverable.

Open the bench
05

What the engine knows

Nineteen distresses, each a reviewed record: the symptom, what separates it from the distresses it is mistaken for, the mechanisms, the causes behind them, the tests that discriminate and the remedies that do and do not address each. Read them here; run them in the bench.

Select a distress 19 records
06

Directed by Dr. Serji N. Amirkhanian

Dr. Serji N. Amirkhanian, Laboratory Director
Dr. Serji N. Amirkhanian
Laboratory Director

A knowledge base is only as good as the engineer who signs off on what goes into it. The laboratory is directed by Dr. Serji N. Amirkhanian, whose work in asphalt technology centres on the three areas this platform goes deepest into: rubber-modified binders, recycled materials, and mixture performance testing.

Review here is not a signature at the end. It is exercised on the records the engine reasons over — so a correction made once is a correction the laboratory cannot subsequently forget.

Deliverables leave marked PENDING APPROVAL until the professional of record signs them. The platform prepares the engineering; the determination is a professional act.

Send us what you already have.

Distress photographs, a mix design, core reports, the contested lot record, or a page of a laboratory report. The first answer you get back will name what is missing before it names anything else — and that is usually the most valuable half hour of the engagement.

Serving state highway agencies, producers, contractors and owners across South Carolina, North Carolina and Georgia.

serji@asphalttechpro.com
1428 Dairy Pearman Road, Anderson, SC 29625