Suplicium Transport LLC told the federal government it runs one truck. Its registration with the Federal Motor Carrier Safety Administration lists a single power unit and a single driver, operating out of Springfield, Illinois. In the same reporting period, roadside inspectors stopped Suplicium’s trucks 801 times in 46 states, on 675 different vehicle identification numbers. No single truck is inspected 801 times across 46 states in a year. The figure on the registration was never a count of equipment. It was the basis for an insurance premium.
Suplicium is one of 32 motor carriers that, taken together, report 38 power units to FMCSA. Inspection records tied to those 32 authorities account for 6,082 unique VINs across 7,505 inspections, with some carriers appearing in as many as 46 states. That works out to roughly 160 vehicles on the road for every truck on the books, and the vehicles do not stay with one company. They circulate.
A migration into the slow lane
The pressure on long-haul Class 8 trucking has built steadily over the past year and a half. English-language proficiency is again an out-of-service violation. The rules on non-domiciled commercial licenses are tightening. The electronic logging mandate closed the hours-of-service gaps that thin-margin carriers once relied on. For an operator built on cheap drivers, light paperwork and a minimum policy, long-haul stopped paying.
Many of those operators moved into hotshot and auto transport. A one-ton pickup pulling a car-hauler attracts fewer inspections than a tractor-trailer, and its insurance costs a fraction of a Class 8 fleet’s. The equipment is also easy to move. A pickup and a gooseneck trailer can be retitled, replated and run under a new limited liability company in a matter of days.
The shift shows up in the equipment. Of roughly 500 vehicles that appear under more than one carrier in this network, 46 percent are Ford and Ram pickups and another 14 percent are Kaufman-style car-hauler trailers. Almost none are heavy freight trucks. The timing tracks the broader auto-haul market. When Jack Cooper, a 97-year-old unionized auto hauler, shut down in early 2025 after losing its Ford and General Motors contracts, the finished-vehicle freight it carried did not disappear. It dispersed down the chain into smaller carriers, and some of those carriers are in this group.
The network is also recent. Twenty-six of the 32 carriers hold DOT numbers issued recently enough to date their authority to 2023 or later, and 21 trace to 2024 and 2025. Their insurance filings follow the same line, with 26 of the most recent policies bound during 2025, most in the spring and summer. This is a wave that formed over the past 18 months and had already generated thousands of inspections before the connections between the companies became visible.
Tenure, or the lack thereof, is where the system is weakest. A newly authorized carrier enters an 18-month new-entrant window during which FMCSA is supposed to run a safety audit before the authority becomes permanent. The problem is arithmetic. The agency processes well over 150,000 new carrier registrations a year and cannot audit a meaningful share of them inside that window. A carrier that draws scrutiny, fails a new entrant review, or has its authority headed for revocation does not have to fight it. It can let that DOT number lapse and file for a fresh one, and the clock starts over with a clean record and another 18 months of runway. The vehicles, the drivers and the people behind them carry forward. Only the identifier changes. A cluster of authorities that are almost all a year or two old is the signature of operators who treat the DOT number as disposable, and who reach for a new one the moment an old one becomes a liability.
The same truck, different doors
The clearest evidence is the shared VIN. One Ford F-350, carrying Illinois plate DRL6327, was inspected 15 times under seven different carrier DOT numbers in 12 states. The most-inspected unit in the set, another Ford on plate P1295664, drew 17 stops across a dozen states. Across the network, single vehicles surface under three, four, five, and six separate authorities. These aren’t driveaway towaway, these are real freight carriers.
The plates swap too. A carrier running its own equipment shows roughly one tag per truck. These carriers show more plates than trucks. Sakara LLC, registered to two power units in Otis Orchards, Washington, accounts for 682 VINs wearing 695 plates. Suplicium shows 681 plates on 675 VINs. Cobra Inc shows 303 plates on 288 VINs. At the level of an individual vehicle, the pattern is insane. One car-hauler turned up under four carriers wearing eight different plates across 14 inspections, most of them near-identical variations on a single Illinois number with a Maine tag mixed in. Others ran on plates reading APPLIED, TEMP, or NOTAG. A trooper at the scale and a toll camera on the interstate both read the plate, not the VIN, so a shifting or temporary tag breaks the link between a violation and the truck that earned it.
The rotation serves a purpose. Spreading inspections and violations across many DOT numbers keeps any single carrier from accumulating the record that triggers an FMCSA intervention. The safety scores stay clean because no one company carries the full history. A bad inspection under one authority is followed by the next inspection under another. After a crash, the vehicle can be tied to whichever entity has the most favorable standing at that moment. In recent attempts to avoid the most active USDOT and FMCSA in history, which is hot on their trail, the answer is to run with no tag at all.
One carrier anchors the group. Auto Haul Express LLC, registered to a single truck in La Grange, Illinois, under officer Olha Kurilovych, appears on 225 of the roughly 500 shared vehicles, close to half the entire crossover set. A second company by the same name, under a different DOT number, is registered to Oleg Shevchenko in Nine Mile Falls, Washington, inside the network’s western cluster. The highest-ratio carrier in the set, Suplicium, lists officer Jamshedjon Ismailov Sr. at a Springfield address while showing a separate officer address in Brooklyn. Its closest equipment partner, Sakara, is run by Ara Sarkisyan, who also controls a second carrier, Rideon LLC, in the same Washington area.
How the companies connect
Secretary Sean Duffy described the problem in nearly the same terms in February, when he said the government cannot allow 200 DOT numbers to trace back to a single post office box.
The ownership ties show up in the people. Anton Mapin, listed on Tsar Transportation in Illinois, also appears on an entity in Connecticut. Gulbakhor Mukhammadieva, on LBS Logistics, also controls a second Illinois carrier. Ara Sarkisyan runs both Sakara and Rideon in eastern Washington. Officer-name matching has limits, since common names collide, but distinctive names recurring across separate authorities are the kind of link a phone-number or address query never catches. The equipment is shared, the formation infrastructure is shared, and a handful of operators turn up again and again behind nominally unrelated companies.
Run as a network rather than a list, the structure is hard to argue with. The carriers that share equipment resolve into a single connected web, not a set of unrelated one-truck firms, and that web breaks down into three tighter operating cells. The same Auto Haul Express that anchors the equipment sits at the center of the network by every measure of connectedness. Tested against a model that randomly reshuffles the same vehicles across the same carriers, the volume of equipment these companies share comes in roughly 20 standard deviations above what chance would produce. A result three standard deviations from the mean is already treated as no coincidence. This is far past that. The shared iron is not an artifact of a busy used-truck market.
The web also has a geography. Of the carriers in and around this network, just under half are registered in Illinois, and another one in seven is in Ohio. The two states together account for nearly 60 percent of the group. Within Illinois, the addresses bunch even tighter, in the Chicago suburbs of Schaumburg, Hoffman Estates, Palatine, Arlington Heights and the registered-agent suites of Springfield. A second pole sits in eastern Washington, around Spokane. When the formation addresses are traced outward to every carrier that shares them, the same suites recur: a single address in Canton, Ohio, ties to 16 carriers, a Spokane suite to a dozen more, a Fargo address to several others in the network. The officers carry the same concentration. The recurring names on these authorities are overwhelmingly Eastern European and Central Asian, the same handful surfacing across nominally unrelated companies and clustered in the same metro areas. That pattern, the same kind of operator forming the same kind of carrier through the same suites in the same cities, is what a list of 236 individual registrations hides and a network map shows.
A number that does not survive
Every carrier files a form called the MCS-150, on which it reports its mileage and truck count. Those two figures are what regulators and insurers use to size a carrier’s exposure, and neither is checked against anything.
In this group, the mileage figure collapses on inspection. Five carriers reported driving exactly one mile for the year. One of them is Auto Haul Express, the same company that appears in 749 inspections across 43 states. Three others reported between 2.9 million and 15.3 million miles on one or two trucks, a figure no truck can produce, since even a hard-run unit covers about 130,000 miles a year. The remaining 21 carriers reported no mileage at all. Of the 32, only one reported a number in a believable range.
The truck count tells the same story in shorthand. One unit, one unit, one unit, against a footprint that crosses the country.
Two insurers, then none
Premium is built on one self-declared truck. The actual exposure is hundreds of vehicles operating under that authority, in dozens of states, driven by people the underwriter never evaluated. The carrier was priced as one truck and is running a fleet.
These carriers are not insured by obscure offshore programs. They are insured by some of the largest names in the business. Of the 32 carriers’ most recent filings, companies in the Progressive group wrote 59 percent and companies in the GEICO group wrote another 28 percent, most through the kind of fast-issue commercial-auto programs that bind coverage off a self-reported application.
The policies are short-lived. Across the group, the filing history shows coverage that rarely lasts a full quarter, with a median of 92 days before cancellation. One policy ran for five days. One was bound and canceled the same day. The same rhythm repeats company to company: bind a minimum policy, run the fleet for a few months, let it lapse, rebind with another insurer. One carrier has moved through nine insurers this way.
Eight entities no longer hold any active authority, having been put out of service by the agency, as the next section lays out. This is the end state that the bind-and-lapse pattern was always heading toward.
The limits sit at the federal floor or just above it, $750,000 for most of the group and $1 million for the rest. A $750,000 policy meets the law on paper. It was priced for one truck and is backed by a fleet that renews its coverage every few months. The MCS-90 endorsement attached to these filings requires the insurer to pay an injured member of the public even when the policy would otherwise deny the claim, so the loss lands on the carrier’s insurer at a multiple of the premium it collected. The insurer recovers it in the only way it can: by raising rates across its entire book. A carrier running 50 trucks and insuring all 50 ends up subsidizing the one that reported a single truck and ran hundreds of trucks. Commercial auto insurance has lost money on underwriting for more than a decade, and fleet-count fraud is part of it, the part no one measures, because the federal government never reconciles the reported number against the road.
A floor set in 1980
The minimum these carriers meet was written for a different industry. The $750,000 liability requirement for general freight was set by the Motor Carrier Act of 1980 and has not been raised since. Adjusted for inflation, it would exceed $2.8 million today. FMCSA itself told Congress that if the figure had merely tracked medical inflation, it would sit above $3 million. The agency opened a rulemaking to raise it in 2014 and abandoned that effort in 2017 after industry opposition from the ATA and OOIDA. OOIDA runs its own Risk Retention Group, so there are obvious reasons why they would not favor a minimum increase. I’d argue that an Association that’s testifying to policymakers on “behalf of drivers” while owning a risk retention group shouldn’t be able to testify that it doesn’t want its minimum coverage limits increased. Call me old-fashioned, but there seems to be a conflict there. Not to mention, it calls into question whether they are testifying in these insurance minimum considerations in the interest of their driver base or in their own interests?
Below the general freight line, the floor drops further. A carrier hauling non-hazardous freight in vehicles rated under 10,001 pounds owes $300,000, and a property carrier whose vehicles fall under that weight is exempt from the federal minimum altogether unless it hauls specific regulated goods. Hotshot work sits right on that boundary. A one-ton dually pulling a loaded car-hauler clears 10,001 pounds easily, which means operators who file at the lower tier or claim the exemption are often misclassified, and no one checks.
The thread running through it all is self-attestation. The carrier sets its own truck count, its own mileage, and, by classifying its own weight, its own insurance tier. Three numbers, all chosen by the party with the most reason to understate them, all accepted on faith by the regulator and the insurer. That made sense when the only way to count a carrier’s trucks was to ask. It does not now.
The crackdown
Enter Sean Duffy, Derek Barrs, the FMCSA investigators and private industry tech. Starting in 2025, this network began to come apart, one authority at a time, and the record of how it came apart is the clearest proof of what it always was.
Almost all of these companies are creatures of the last few years, and the registration records show the front door has never stopped opening. A handful trace to 2021 and 2022. Seven were granted authority in 2024. The single largest wave, 17 of the 32, came in 2025, even as chameleon carriers were drawing national scrutiny. One more, Golla, registered in January 2026. The carriers the agency is shutting down now were, for the most part, admitted to the road over the same stretch of months it spent building the case to remove them. The pattern that should have stopped them was not being read at the point of entry, so they scaled fast, ran up inspections and crashes far beyond what a one-truck operation should generate, and drew a hard audit only after the record was built. The systematic shutdown has intensified over the past year, with the heaviest impact in recent months. Enforcement is catching and closing them in near real time. The registration system is still letting them in.
On November 28, 2025, FMCSA put Auto Haul Express out of service. The order followed a failed new-entrant audit and a denial of access; the carrier would not produce the records the agency asked for. Over the next five months, twelve more of the core carriers fell the same way, nearly all of them for refusing an audit or refusing contact. Sakara went out on December 7. Central Transit’s authority was revoked effective December 8. MLS Capital fell on February 2, 2026; Lipchenko on February 7; DMLCSL and ONEPOINT on March 14; Hauldex on March 16; MTMFO’s authority was revoked effective March 25; Novaline went out on April 6; LCZW on April 24; Golla on April 27; and Shapoval on April 29. Thirteen authorities at the center of the cluster, gone in five months and a day. For those claiming enforcement is not acting, you have no idea how wrong you are until right now.
That is enforcement at work, and it deserves recognition. The out-of-service tool that took these carriers down is not new, and the loophole they worked sat open for years under administrations of both parties. What changed is the posture. The agency has made unmasking chameleon carriers a stated priority and has paired it with a push to restore the requirement that a carrier keep a real, inspectable place of business. The cluster coming apart on a five-month timeline is what that posture looks like in the data.
The posture now comes with machinery. In late 2025, the agency began phasing in Motus, a rebuilt registration system that adds identity verification and business validation at the front door, the exact point where a reincarnated carrier has always slipped through. It has moved to enforce the principal-place-of-business rule and announced 40 additional investigators. In February 2026, bipartisan legislation, the Safety and Accountability in Freight Enforcement Act, directed the FMCSA to develop and test an automated tool to flag chameleon applications during registration, codifying a data-driven detection approach the government had once funded more than a decade ago and then let idle. The thirteen authorities in this cluster did not come apart on their own. They came apart while the agency was rebuilding the door they had walked through. Administrator Derek Barrs, Secretary Sean Duffy and the Trump Administration are not only shutting the door, but they’re cleaning out the closet at the same time with minimal resources.
Killing the authority did not take the trucks off the road. It moved them. When Auto Haul Express went out on November 28, its equipment did not park. Within 48 hours, the same vehicle identification numbers were turning up under other carriers in the group. MLS Capital and Central Transit inspected Auto Haul’s trucks on November 29, DMLCSL and Jay Torres on November 30. Over the following weeks, 92 of Auto Haul’s vehicles surfaced under Hauldex, 84 under MTMFO, 72 under DMLCSL, with dozens more spread across Lipchenko, LCZW, Golla, Auto Titi, and Shapoval. The hub authority was dead. Its fleet kept rolling under new numbers.
When Sakara went out nine days later, the pattern repeated at a larger scale. The day after its December 7 shutdown, 236 of Sakara’s VINs began appearing under Hauldex. Another 102 went to Golla, 62 to Richroad, 51 to MTMFO. The trucks did not change. The authority on the paperwork and the door markings did.
As the agency worked down the list, the carriers that had caught the equipment became targets in turn, and the iron moved again. Hauldex absorbed hundreds of vehicles from Auto Haul and Sakara over the winter, and then on March 16, Hauldex itself was put out of service. The next day, 51 of the vehicles that had been running shifted to Golla. Golla’s authority was revoked six weeks later, on April 27. The same trucks had now outlived four separate authorities and were looking for a fifth. Across the cluster, more than 1,300 vehicles passed through carriers that were each, in turn, shut down, equipment walking down a line of shells, every one of them dying behind it.
The newest authorities tell the rest of the story. KDN Transport Group, formed in Charleston, Illinois, and NHX Logistics, formed in Plano, Illinois, were both brand-new carriers in early 2026, each carrying a fresh million-dollar policy. As the older shells were being killed off through the winter and spring, these two were catching their equipment. KDN picked up vehicles from five different out-of-service carriers, with its first inspections dated March 17 through March 25, the most recent in the entire record. NHX took 48 vehicles from the freshly closed MLS Capital. The oldest iron in the network was landing on the youngest authorities, the ones still too new to have drawn an audit.
The plates make it concrete and current. One Illinois tag, 209267F, was read by inspectors on trucks operating under Jay Torres, MTMFO and Golla, three nominally separate carriers, the last reading dated March 24, 2026. A second Illinois plate moved across the same three companies. MTMFO and KDN share more than a dozen tags between them, the most recent seen on March 25. These are the same physical plates moving between active carriers inside the past several weeks, while the enforcement that killed the parent companies was still underway.
That is how fast “chameleon activity” occurs. The agency can revoke an authority, and it is doing so more quickly and deliberately than before. Revocation acts on the number, and the number is the one thing in this network that is disposable. The trucks, the plates, and the people behind them re-form under the next clean authority, and the new-entrant clock starts over. It reforms fast. The same equipment moved within 48 hours of each shutdown, and new authorities were being created even as the old ones were killed. KDN and NHX were formed in January 2026, in the middle of the very enforcement wave that was dismantling the carriers whose equipment they would inherit weeks later. That is the network showing its hand, how quickly it can change names and shift assets to stay a step ahead of the agency and its investigators. Until the reported fleet count is reconciled against the inspection record at the moment a carrier registers, the door the old authorities walked out of stays open for the new ones to walk back in.
What the inspections were finding
Behind the paperwork, the same network incurred more than 17,000 violations in its inspection record. The single most-cited violation across these carriers, appearing at more than half of them and recorded 775 times, is operating a commercial vehicle without a valid commercial driver’s license. In nearly every instance, the driver was placed out of service on the spot. A separate citation for operating without a CDL adds hundreds more. Alongside them sit 380 citations for drivers who could not satisfy the English-language proficiency requirement, again overwhelmingly resulting in an out-of-service order.
The equipment failures track the licensing failures. Hundreds of out-of-service brake violations, missing or inoperable breakaway systems on the trailers, defective brakes exceeding a fifth of the braking system, bald tires below the legal tread depth, and tires leaking or carrying weight past their rated limit. These are the conditions that turn a loaded car hauler into something that cannot stop or stay in its lane. Read together, the violation record describes the predictable result of the fraud on the registration form: when the entity exists only to carry an insurance filing and the trucks are run hard under disposable authorities, no one is checking whether the driver is licensed or whether the brakes work.
The crash record is where it lands. The carriers in and around this network appear in more than 3,000 crashes. The toll is heavily concentrated in the larger fleets the network’s equipment touches, which is its own warning, but even confined to the small one- and two-truck shells at the core, the carriers reporting a single vehicle while running dozens, the record holds hundreds of crashes, dozens of fatalities and hundreds of injuries. Manas Express, which reported two trucks and put more than 360 distinct vehicles through inspections, appears in 240 crashes with five deaths and remains listed as active. This is the public-safety cost of a self-declared fleet count and an authority that can be replaced in an afternoon.
A death at the end of the chain
The cost is not only financial. I have served as an expert witness on multiple fatal interstate crashes that began in exactly this part of the industry, with a hotshot operator. In one case, a pickup pulling a car hauler and a driver who held only a Russian commercial license. The load had been brokered and re-brokered four times through enough hands that establishing who controlled the driver, who dispatched the freight and who was responsible for confirming the driver belonged behind the wheel was nearly impossible. Someone died. The driver had no business operating a commercial vehicle, and the carrier’s reported fleet size bore no relation to what it was actually running. The killer piece of this case specifically wasn’t even the Russian CDL; it was the flashlights mounted to the rear of the trailer serving as marker lamps that ultimately killed a man who didn’t see the trailer and hit it at night.
That is what the diluted safety score, the lapsed minimum policy, and the layered brokerage add up to on a dark interstate. The structure is built to keep the responsible party out of reach until a death forces someone to trace it back.
The brokers are exposed now
For years, the layering served as both a legal and a practical shield. When a crash led back to a broker who had placed the load with an unsafe carrier, the broker’s defense was federal preemption: the Federal Aviation Administration Authorization Act of 1994 bars state laws related to a broker’s services, and most courts read that to knock out state negligent-selection claims before they reached a jury. A broker who booked a ghost carrier could argue it was not the law’s business how it chose.
That defense is gone. On May 14, 2026, in Montgomery v. Caribe Transport II, LLC, the Supreme Court ruled unanimously that the FAAAA does not preempt state-law claims that a broker negligently selected an unsafe motor carrier. The Court placed those claims inside the statute’s safety exception, which preserves a state’s authority over motor-vehicle safety, reasoning that a broker’s choice of carrier directly determines which trucks and drivers end up on the road. The decision resolved a long-running split among the circuits and applies immediately to pending cases and future ones alike. A broker that books a load with a carrier reporting one truck and running hundreds, with a safety record built on unlicensed drivers and out-of-service brakes, can now be made to answer for that choice in front of a jury.
The catch is that being liable and being able to pay are not the same thing, and freight brokers sit on the wrong side of that gap. A carrier has to carry at least $750,000 in liability coverage, thin as that floor is. A broker carries no liability insurance requirement at all. The only federal financial requirement for a broker is a $75,000 surety bond, which exists to cover unpaid carrier and shipper claims, not to pay out a wrongful-death judgment. Most brokerages are small, and a large share of the market is handled by independent agents operating under their own authority with few assets behind them. So the broker who is most likely to place freight with a ghost carrier, the one not vetting at all, is frequently the one with nothing to collect against when the negligent-selection claim succeeds. Justice Kavanaugh, concurring, noted that the litigation and insurance costs of the decision will be real even for brokers who ultimately win, and that those costs will work their way through the economy. The honest, well-capitalized broker will buy contingent liability coverage and tighten its vetting. The fly-by-night agent will carry the bond, book the cheap carrier and have little to surrender when the case is traced back. Accountability, in other words, lands hardest where there is something to reach, which is not always where the freight was actually placed.
This is the same hole that runs through the entire scheme. A carrier insured at a 1980 minimum, or not insured at all, hauling under an authority that can be discarded and reformed, booked by a broker with a bond and no assets. At every link, the financial responsibility that is supposed to stand behind a loaded truck on a public road has been engineered down to the smallest number the rules allow, or below it.
The fix is a query
The data needed to catch this already exists inside FMCSA. Every inspection record carries the VIN, the plate, the carrier’s DOT number, the date, and the location. Comparing a carrier’s reported truck count with the number of unique VINs that appear in its inspections is a database query. We built one. It runs against millions of records in minutes and finds these groups.
A ratio of 10 vehicles to one reported truck should prompt a review. A ratio of 160 to one should prompt an investigation. An insurer should not be able to bind a one-truck policy for a carrier that already has hundreds of VINs in the federal inspection record without first seeing that record.
There is movement, and it is moving fast. FMCSA Administrator Derek Barrs said in February 2026 that unmasking chameleon carriers was a priority and that the agency was restoring enforcement of the requirement that a carrier have a real principal place of business. The reforms announced alongside it would require carriers to keep a physical location where records can be inspected within 48 hours, move English-proficiency violations from out-of-service status to license revocation, and add new rules on suspensions, new-entrant vetting, and broker testing. Motus registration system is now live with identity verification and business validation aimed at the shell entities and ghost offices this network runs on, and Congress has moved to require an automated chameleon-detection tool at the point of registration. Those measures attack the formation mills and the mailbox addresses. They do not yet reconcile the reported truck count with the inspection record, the single step that would expose the fleet-size fraud at the center of this.
The same signals that exposed this cluster, shared VINs, migrating plates, and common formation addresses, are now being read by a layer of tools the agency has brought into its own work. I have worked with the FMCSA on multiple cases like this for some time with my own system, and the FMCSA has also adopted the use of GenLogs, which runs a sensor network that reads plates off the road, and Bluewire, a carrier-risk scoring system, both of which are in active procurement use against the same or similar problems. Paired with the rebuilt Motus registration system, verifying identity and business legitimacy at the front door, the federal government and the private market are, for the first time, reading the same data and arriving at the same carriers. The capability is no longer the missing piece. The will to run it continuously and to act on what it returns has changed.
The floor has to move as well. A bill introduced in the House in April 2026, the Fair Compensation for Truck Crash Victims Act, would raise the minimum to $5 million and index it to inflation so it cannot calcify again for 40 years. The objection will be that it burdens small carriers. The carriers it would actually burden are the ones in this story, reporting one truck and running hundreds, and they are the ones it should. The honest operator already carries at least $1 million in coverage. The fraud carries the minimum, lets it lapse, and leaves the public and every law-abiding carrier to cover the gap.
The agency knows it has taken on more than it can comfortably handle. Administrator Barrs has said as much, that they bit off more than they could chew, and they are going to keep on chewing anyway. In thirteen shuttered authorities and a year of audits, a list of refusers was whittled down; that is what the record looks like. The other side is fast, sprinting to stay one identity ahead. For the first time in my years in this industry, the distance between a fraud and the enforcement that catches it is closing toward real time. Sometimes that level of focus, ambition, and drive is all it takes.
The post Inside hotshot trucking’s ghost fleets appeared first on FreightWaves.
