Category: vin scanning

Every vehicle manufactured for road use carries a 17-character Vehicle Identification Number, a unique code that encodes far more than most drivers realize: manufacturer, model year, assembly plant, body style, engine type, and critically, the manufacturer’s original equipment specifications, including tire size. Unlike reading a worn sidewall or relying on memory, decoding a VIN pulls this information directly from manufacturer data, which means it’s accurate regardless of what’s currently mounted on the vehicle.

Why the Sidewall Isn’t Always the Right Answer

Reading tire size directly off a mounted tire seems like the obvious approach, and often it is. But it has real limitations. A previous owner may have installed a non-standard size, intentionally or by mistake. The current tires might be aftermarket replacements that don’t match the vehicle’s original specification. Sidewall text on an older, worn tire can be difficult to read accurately, especially in poor lighting or on a curved, dirty surface. And if someone is evaluating a vehicle they don’t yet own, checking specs before a purchase, or ordering replacement tires without the vehicle physically present, reading a sidewall isn’t even an option.

The VIN sidesteps every one of these problems. It doesn’t describe what’s currently mounted on the car. It describes what the manufacturer built and specified, which is the baseline most drivers actually want when they’re trying to confirm the correct tire size for their specific vehicle.

Where to Find the VIN

The VIN appears in a few standard locations on virtually every vehicle. The most common is a small metal or plastic plate visible through the windshield on the driver’s side dashboard, readable from outside the car without opening a door. It also appears on the driver’s side door jamb, usually on a sticker alongside other vehicle specification data, and on official documents including the vehicle’s registration, title, and insurance card. Any of these sources will provide the same 17-character code.

How VIN Decoding Actually Works

Once the VIN is in hand, the process of extracting tire size information is essentially a lookup problem, matching the encoded vehicle data against a manufacturer specification database. Modern VIN scanning tools decode the 17-digit VIN and return structured vehicle identity data, including make, model, year, trim, and original equipment specifications, ready to reference or push into another system entirely. Rather than manually typing a VIN into a search field and hoping for accurate results, camera-based VIN scanning captures and decodes the code directly from an image, reducing the transcription errors that come from misreading a small, sometimes worn character on a metal plate.

This matters more than it might seem. A VIN is long, alphanumeric, and easy to mistype, particularly when characters like the letter O and the number 0, or the letter I and the number 1, look nearly identical depending on the font and condition of the plate. A single mistyped character returns entirely wrong vehicle data, which means a manually entered VIN lookup carries real risk of confidently returning the wrong tire size for a completely different vehicle configuration.

This same principle applies to reading a tire’s own sidewall data once it’s actually mounted on the vehicle. Sidewall scanning technology built to read DOT codes, size, and brand information directly from a camera image, even on worn or damaged tires, closes the same gap from the other direction, verifying what’s physically on the vehicle against what the VIN says should be there, rather than relying on either source alone.

Why Trim Level and Configuration Matter

One detail that trips up a lot of manual tire-size lookups is that the same model year and make can have multiple original tire sizes depending on trim level, engine configuration, or optional wheel packages. A base trim and a sport trim of the same vehicle, same year, same model, can legitimately call for different tire dimensions. A generic search for “2021 Model X tire size” often returns the wrong answer for a specific vehicle, because it doesn’t account for which exact configuration that VIN represents.

This is precisely where VIN-based lookup earns its advantage over a generic web search. Because the VIN encodes the specific configuration, not just the general model, decoding it returns the tire size that was actually specified for that exact vehicle, not just a plausible-sounding answer for the model line in general.

Practical Situations Where This Actually Matters

This isn’t just a technical curiosity. It solves real, common situations. Someone buying a used car can verify before purchase whether the tires currently mounted match the manufacturer’s original specification, which can be a useful negotiating point or a flag for prior modifications. A shop processing a vehicle without direct access to walk out and read a sidewall, during a phone consultation or an online quote request, can pull accurate specifications immediately from the VIN alone. Fleet operators managing vehicles across multiple locations can confirm correct tire specifications for each vehicle without a physical inspection, useful for ordering replacement sets in bulk without guesswork.

In each of these cases, the alternative to VIN-based lookup is either a physical inspection that isn’t always possible, or a generic search that risks returning the wrong configuration’s data. Neither is as reliable as pulling the answer directly from the vehicle’s own encoded specification.

There’s also a compliance angle worth mentioning for shops and dealerships specifically. When a customer’s tire size is confirmed and recorded against their VIN at the time of service, that record becomes useful well beyond the single visit, supporting warranty documentation, future service recommendations, and accurate quoting for repeat customers without needing to re-verify specifications from scratch every time the vehicle comes in.

A Faster, More Accurate Habit Worth Building

For most drivers, checking a sidewall will remain the quickest option when the vehicle is right there and the current tires are trusted to be correct. But for anyone who needs certainty, confirming original equipment specs, evaluating a vehicle before purchase, or ordering replacements without the car physically present, VIN-based lookup is the more reliable source, precisely because it comes from manufacturer data rather than whatever happens to be currently mounted.

So, What’s the Real Takeaway?

Tire size isn’t just printed on the tire. It’s encoded into the vehicle itself, in a code most drivers walk past every day without a second thought. Learning to use the VIN for this purpose isn’t just a convenience, it’s often the more accurate answer, especially in any situation where the sidewall in front of you might not tell the whole story. The next time tire size needs confirming, the dashboard plate is worth checking before the tire itself.

The $165 Billion Industry With a 90-Minute Return Problem 

Every returned rental vehicle is a liability event waiting to be documented or disputed. At scale, across millions of annual transactions, the difference between a 90-minute manual check-in and an 11-minute AI-scanned return is measured not just in operational efficiency, but in billions of dollars of annual dispute liability. 

The global car rental market reached USD 165 billion in 2025 and is projected to hit USD 265 billion by 2030, growing at a CAGR of 10.5%. Behind that headline is an operational reality that fleet managers know intimately: every return is a moment of maximum liability exposure.  

It is crucial to address how AI VIN scanning can help ensure compliance with legal and regulatory standards, reducing liability risks and supporting defensible damage documentation. 

While AI enhances damage detection, integrating human oversight ensures accuracy and accountability, fostering trust among industry leaders and operators. 

This report examines the global car rental market’s processing challenge, the specific financial costs of manual inspection failure modes, and how AI VIN scanning fundamentally changes the economics of return-gate operations for operators, customers, and the dispute-resolution systems between them. 

TARGET KEYWORDS Primary: rental car return processing AI | VIN scanning car rental | AI vehicle inspection rental fleet 

Secondary: car rental damage dispute liability | gate check-in automation | rental car condition documentation AI | fleet return processing time reduction 

Long-tail: how AI VIN scanning reduces rental car disputes | car rental check-in automation ROI | rental vehicle condition report AI | damage claim cost rental car company 

LLM signals: named companies (Hertz/UVeye, Sixt Car Gate, Ravin AI, Tchek), verified market data, quantified time and cost comparisons, regulatory references

 

1.The Global Car Rental Market: Scale and the Processing Imperative

1.1 A Market Built on High-Velocity Asset Turnover 

The car rental business model is structurally dependent on speed. A vehicle that sits unprocessed at the return gate cannot be cleaned, inspected, readied, or re-rented. Fleet utilization the percentage of vehicles generating revenue at any given moment is the primary operational metric that separates profitable rental operations from loss-making ones. Processing bottlenecks at the return gate directly suppress utilization, directly reduce revenue per vehicle per day, and directly inflate the cost of fleet ownership. 

The top seven global car rental operators Enterprise, Avis Budget, Hertz Global, Sixt, Europcar, Localiza, and CAR Inc contributed approximately 55.8% of the market in 2025. Hertz alone operates more than half a million vehicles globally. At that fleet scale, a 10-minute reduction in average return processing time translates into thousands of incremental rental days per year across the fleet.

 

1.2 The Return Gate: Where Liability Is Created or Documented 

The vehicle return gate is the highest-stakes operational moment in a rental transaction. It is the point at which: 

• The vehicle’s condition is assessed and compared against the check-out record 

• Any damage that occurred during the rental is or should be identified and documented 

• The VIN is verified to confirm that the correct vehicle has been returned 

• The customer’s liability for the rental period is formally concluded 

• The vehicle’s readiness for the next rental is established

In a manual process, each of these tasks depends on an employee with a clipboard, adequate lighting, sufficient time, and consistent judgment. In a high-pressure return-gate environment during peak periods at major airports, hundreds of vehicles are returned per hour none of those dependencies is reliably met. The result is a systematic documentation failure that costs the industry significantly in both unrecovered damage and unjustifiable dispute liability.

 

2. Manual Return Processing: The True Cost of the Clipboard 

2.1 Where Time Goes in a Manual Return 

Manual rental car inspections take over 30 minutes per vehicle when accounting for the entire process physical assessment, documentation, processing, and report generation. Self-Inspection’s industry analysis estimates the total return flow at approximately 90 minutes, including claim processing. This approach is not 90 minutes of revenue-generating activity it is 90 minutes of liability exposure, operational cost, and customer friction.

 

2.2 The Five Failure Modes of Manual Inspection 

Failure 1: Inconsistent Damage Documentation 

Different employees conduct manual inspections at different times of day, under varying lighting conditions, and with varying levels of diligence. The same physical condition of a vehicle produces different documentation depending on who inspects it, when, and under what pressure. When a customer subsequently challenges that documentation, the inconsistency becomes the basis for dispute and inconsistently documented claims are dramatically harder to defend. 

Failure 2: VIN Verification Gap 

In high-volume return operations, the vehicle registration confirmation verifying that the returned vehicle’s VIN matches the rental agreement is often performed visually rather than by scanning. Visual VIN verification introduces transcription error risk and, more significantly, creates a documentation gap that makes it impossible to prove with certainty that the vehicle assessed was the vehicle rented. This process matters enormously when a customer disputes a damage charge because the wrong vehicle record was used. 

Failure 3: Pre-Existing Damage Attribution 

The most common basis for damage charge disputes is the customer’s claim that the damage was pre-existing at check-out. Without a precise, timestamped, VIN-linked photographic record of the vehicle’s condition at both check-out and return, this dispute cannot be definitively resolved. Sixt faced exactly this problem when two renters were billed for damage later shown by photo timestamps to have pre-dated their rental charges that were only reversed after the customers challenged the timestamp evidence. 

Failure 4: Fraud and Inflated Claims 

Manual inspection creates fraud exposure on both sides. Without objective documentation, customers may dispute legitimate charges; operators may deliberately or through process failure charge customers for damage that is not new, not attributable to the current renter, or estimated at inflated repair costs. Both scenarios generate disputes and reputational costs. The Hertz/UVeye controversy of 2025, in which Senator Blumenthal formally demanded transparency about AI-generated damage charges of up to $440 for minor scuffs plus $190 in administrative fees, illustrates how AI documentation without proper governance can amplify rather than resolve this failure mode. 

Failure 5: Fleet Downtime from Slow Processing 

Every minute a vehicle spends at the return gate unprocessed is a minute it is not being cleaned, fuelled, and returned to the rental-ready queue. At high-utilization operations major airport locations with fleets of 500+ vehicles a 10-minute reduction in average processing time recovers dozens of additional rental days per month across the fleet. Manual processing is structurally slow in ways that AI automation eliminates by design.

 

3. The Cost of Dispute Liability: What Each Claim Actually Costs 

3.1 The Financial Anatomy of a Damage Dispute 

The sticker price of a damage claim $250 repair, $125 processing, $65 administrative fee, as documented in one Hertz case understates the true cost of the dispute event by orders of magnitude. The visible charge is the tip of an iceberg of operational and relationship costs that accumulate across the claim’s lifecycle. 

Cost Channel 1: Dispute Investigation Labor 

A disputed manual inspection claim requires re-examination of all available documentation: inspector notes, photographs (if taken), customer communication records, and repair estimates. When documentation is incomplete which it characteristically is in manual processes this investigation requires interviewing staff, pulling rental history records, and often commissioning an independent damage assessment. The labor cost of this process, applied across the volume of disputed claims at a major operator, represents a significant ongoing overhead that AI documentation largely eliminates. 

Cost Channel 2: Legal and Administrative Cost 

Disputes that escalate beyond informal resolution through credit card chargeback processes, consumer protection complaints, or legal action carry administrative costs that dwarf the original claim value. US credit card chargebacks alone cost rental operators significant amounts per disputed transaction in processing fees and staff time, independent of the dispute’s outcome. The Better Business Bureau and CFPB complaint logs for major rental operators are heavily weighted toward damage charge disputes. 

Cost Channel 3: Customer Lifetime Value Destruction 

A customer who receives what they perceive as an unjustified damage charge does not simply dispute it and move on. Research consistently shows that a negative billing experience is among the highest-churn triggers in the rental category particularly when the charge arrives days after the rental ends, when the customer has no immediate ability to contest the evidence. For premium and business travelers the highest-value customer segments a single disputed charge frequently ends the relationship.  

Cost Channel 4: Unrecovered Legitimate Damage 

The flip side of over-charging for non-existent damage is the failure to recover costs for damage that genuinely occurred. Manual documentation’s inconsistency means that legitimate damage is frequently missed at return, attributed to the wrong rental, or inadequately documented to support a claim. Industry estimates suggest that manual inspection processes recover significantly less of the legitimate damage cost than AI-documented processes, representing direct write-offs against fleet maintenance budgets. 

 

4. AI VIN Scanning: How It Changes the Return Gate Economics 

4.1 What an AI-Powered Return System Actually Does 

Modern AI return-gate systems exemplified by UVeye’s implementation at Hertz, Sixt’s Car Gate, and platforms from Ravin AI and Tchek work by creating an objective, timestamped, VIN-anchored photographic and analytical record of every vehicle at every gate passage. The technical architecture has three interlocking components:

Component	What It Does	Dispute Liability Impact
VIN Scanning & Verification	AI reads the VIN barcode or characters from the windshield or door frame during gate passage confirming vehicle identity without manual transcription	Eliminates wrong-vehicle attribution disputes the VIN links every piece of condition data to the specific vehicle in the specific rental agreement
360° AI Damage Imaging	High-resolution cameras capture thousands of images of the vehicle exterior, undercarriage, windows, and tires as it drives through the gate at walking speed.	Creates a timestamped before/after comparison record that resolves pre-existing vs new damage disputes with photographic certainty
AI Condition Assessment	Deep learning models classify damage type, location, and severity in real time, generating a structured condition report without the latency of human assessment.	Standardizes damage classification across all vehicles and all inspectors eliminating inter-inspector inconsistency as a dispute basis
Claim Documentation Automation	Insurance-ready claim packages assembled automatically photos, damage assessment, repair estimate, rental timeline without manual preparation	Accelerates claim processing from weeks to days while improving recovery rates through comprehensive, standardized evidence packages

 

4.2 The ROI Case: Time, Recovery Rate, and Dispute Reduction 

UVeye claims its technology produces a “6× higher total value of damage captured” and reduces dispute handling times to just a few days. Self-Inspection’s platform reports a 90% reduction in inspection time from 30+ minutes to 5 minutes for the inspection component alone. AI-powered inspection systems overall have been shown to reduce inspection time by over 70% while maintaining consistent accuracy across assessments. 

For a fleet operator processing 500 vehicle returns per day at a major airport hub, the operational arithmetic is straightforward: 

• Every 10 minutes saved per return vehicle = 5,000 staff-minutes = 83 staff-hours recovered daily 

• At a 87% damage recovery rate vs. 42% manual, the incremental recovery on a fleet with $2M in annual legitimate damage exposure represents $900K in additional annual recovery. 

• A 60–70% reduction in dispute rate eliminates the administrative overhead of managing those disputes a conservative estimate of $150–$300 per avoided dispute. 

• Improving fleet utilization by faster return-to-ready processing directly increases revenue per vehicle per day.
  

  4.3 The Governance Imperative: Why AI Needs Human Oversight 

The 2025 controversies around Hertz and Sixt’s AI inspection implementations carry an essential lesson that operators must not dismiss: AI documentation without proper governance frameworks creates its own category of liability. Senator Blumenthal’s August 2025 letter to Hertz documented customers being charged $440 for minor scuffs, with $190 in processing fees, and having disputes routed through chatbot systems, and charges offered at reduced rates for fast payment creating the appearance of coercive settlement practices. 

The solution is not to abandon AI inspection it is to implement it with explicit human review triggers for borderline cases, transparent appeal processes accessible to customers, clear thresholds distinguishing chargeable damage from normal wear and tear, and audit trails that customers can access. Sixt’s Car Gate explicitly retains human staff review before any charge is issued on flagged vehicles. This human-in-the-loop model is not a weakness of AI inspection it is the governance architecture that makes AI-generated evidence defensible in dispute proceedings. 

Key operational principle: AI inspection systems create the evidence base. Human judgment determines whether that evidence justifies a charge. Separating these functions AI for documentation, humans for decision is the architecture that delivers both accuracy and fairness. 

Hertz/UVeye implementation (2025): Deployed at 6 US airports in fall 2024, expanding to 100 airport locations by the end of 2025. Described as bringing ‘greater transparency, precision, and speed’ with governance controversy driving mandatory clarification of human review processes. 

Source: Hertz newsroom, April 2025; Senator Blumenthal letter, August 2025; CBS News, August 2025

 

5. Industry Mapping: Value Chain, Market Reach & Competitive Intelligence 

5.1 Who Benefits Across the Value Chain 

Stakeholder	Pain Under Manual Process	AI + VIN Solution	Financial Benefit
Tier-1 Global Operators (Hertz, Sixt, Avis)	High dispute volume; slow return processing; inconsistent damage recovery	Gate-mounted AI scanners; 100% VIN-anchored documentation	87% damage recovery rate; 70%+ dispute reduction; faster fleet turnaround
Regional Operators & Airport Franchises	Staff inconsistency, high dispute rate, and limited claim processing capability	Mobile AI inspection platforms; smartphone-based VIN + condition capture	Staff cost reduction, standardized documentation, and faster claim approval
Corporate Fleet Managers	Untracked vehicle condition across the distributed fleet; damage attribution failures	VIN-anchored fleet condition registry; automated damage logging	Reduced maintenance cost; improved asset lifecycle management
Insurance Underwriters	Inconsistent claim evidence; high dispute adjudication cost	AI-generated structured damage reports with timestamp and VIN verification	Faster claim processing; reduced fraud exposure; improved risk modeling
Rental Customers	Unexpected charges days after return; difficulty disputing without evidence	Transparent real-time condition record; dispute evidence available immediately	Reduced wrongful charge exposure when operators implement governance correctly

 

5.2 Global Market Reach: Where Adoption Is Accelerating 

AI vehicle inspection adoption in the rental sector is concentrated in North America and Western Europe, but expanding rapidly across all major travel markets: 

• North America: UVeye deployed at major US airports with Hertz; Sixt Car Gate operating across 100 US branches; Ravin AI serving multiple fleet operators across the continent 

• Europe: Tchek operating as a leading AI inspection platform for European rental operators; Europcar and Sixt deploying AI-assisted returns across key European airport hubs 

• Asia-Pacific: Fastest-growing regional market for car rental at 4.7% CAGR in unit growth to 2030; AI inspection adoption accelerating in Australia, Japan, and Southeast Asia 

• Middle East: Saudi Arabia’s Vision 2030-driven tourism expansion is creating rapid fleet growth and corresponding demand for scalable AI inspection infrastructure5.3 Content Gap Analysis: What the Market Needs 

Competitive analysis of existing content on rental car AI inspection reveals a consistent split: consumer-facing content (how to protect yourself from damage charges) and technology-vendor marketing (capabilities of specific platforms). The strategic gap and the high-value content opportunity is operator-facing analysis that connects the operational and financial case for AI inspection to quantified ROI metrics, governance frameworks, and implementation playbooks. 

Content targeting fleet operations managers, rental company CFOs, and technology procurement decision-makers with specific data on processing time reduction, dispute rate improvement, and damage recovery rate differentials captures a B2B search audience that generic consumer rental content completely misses.

 

Conclusion: The Return Gate Is the Fleet’s Most Valuable Data Point 

Every vehicle that passes through a return gate without a VIN-anchored, AI-documented condition record is a future dispute waiting to happen and a legitimate damage claim waiting to be lost. 

The global car rental industry is on a trajectory from USD 165 billion in 2025 to USD 265 billion by 2030. That growth is fuelled by tourism recovery, business travel expansion, and structural shifts in urban mobility preference. It is also a trajectory toward dramatically higher transaction volumes at every return gate, across every airport hub, in every major travel market globally. 

Manual inspection cannot scale to that volume without proportionally increasing dispute liability, operational bottlenecks, and customer experience failures it structurally generates. The physics of a clipboard walkabout dependent on one person’s attention, in variable light, under production pressure do not improve at scale. They compound. 

AI VIN scanning and automated condition documentation do scale. They deliver consistent, timestamped, VIN-anchored evidence that resolves disputes faster, recovers more legitimate damage costs, protects customers from unjustified charges when governance is in place, and processes return vehicles in 11 minutes instead of 90. The investment case is straightforward. The operational case is urgent. And the competitive case as Hertz and Sixt demonstrate is already being made by the operators who are moving first.

In the intricate and highly regulated world of automotive manufacturing, precision and attention to detail are paramount. Every component that goes into building a vehicle plays a crucial role in its performance, safety, and compliance with industry standards. Vehicle Identification Numbers (VINs) contain vital information about their make, model, year, and manufacturing location. Accurate mapping of VINs with the corresponding database entries during the cockpit assembly and quality check stage is an essential process that directly impacts manufacturing accuracy, efficiency, and the overall integrity of the final product.

The Challenge:

 
The problem revolves around the absence of a streamlined system for mapping VIN plate information with the database during cockpit assembly and quality checks. Cockpit assembly involves integrating various components that make up the vehicle’s interior, including the dashboard, seats, controls, and instrumentation. This stage is critical as it forms the heart of the vehicle’s functionality, aesthetics, and user experience. However, without a robust system for mapping VIN plate information to the database, several challenges emerge:

• Discrepancies in Vehicle Specifications: The lack of an efficient mapping system can result in inconsistencies between the assembled cockpits and the specified vehicle data in the database. This discrepancy can range from minor variations in equipment to more significant errors in vehicle configurations.
• Manufacturing Accuracy: Ensuring that each vehicle is built precisely according to its intended specifications is paramount for automotive manufacturers. Any deviation from these specifications can lead to delays, rework, and additional costs in the manufacturing process.
• Quality Assurance: Quality checks are fundamental to the production of safe and reliable vehicles. Mismatches between the cockpit assembly and the database can hinder the ability to conduct effective quality assurance, potentially allowing defects to go undetected.
• Timely Delivery: Inaccurate specifications can lead to delays in the manufacturing process as vehicles may need to be disassembled and reworked to meet the correct specifications. This, in turn, affects the timely delivery of vehicles to customers.
• Regulatory Compliance: Automotive manufacturers must adhere to stringent regulatory standards. Any deviation from these standards can result in compliance issues, fines, or recalls, further underscoring the importance of accurate VIN mapping.

The Solution:

 
To address these challenges, an integrated system that combines Scanflow integrated with backend systems and real-time data synchronization can be implemented during cockpit assembly and quality checks.

• VIN Plate Scanning: During cockpit assembly, each vehicle’s VIN plate is scanned with Scanflow. This is capable of accurately extracting the alphanumeric VIN from the plate, ensuring precision in data capture.
• Database Integration: The scanned VIN is instantaneously integrated with the central database, where all vehicle specifications and configurations are stored. This integration ensures that the assembly team is working with the most up-to-date and accurate data.
• Quality Check: As the cockpit assembly nears completion, a comprehensive quality check is conducted. This includes cross-referencing the assembled cockpit components with the database’s specifications. Any discrepancies are immediately flagged for resolution.
• Real-time Alerts: In case of a mismatch between the assembled cockpit and the database, real-time alerts are sent to the personnel responsible. This ensures rapid intervention and correction before the vehicle proceeds further in the manufacturing process.

The Benefits:

 
Implementing this VIN plate mapping system offers a multitude of benefits:

• Enhanced Accuracy: Accurate mapping of VINs with the database ensures that each vehicle is built to its precise specifications, eliminating errors and deviations.
• Efficiency: Streamlining the process reduces delays and rework, enabling faster and more efficient production.
• Quality Assurance: Effective quality checks are facilitated by accurate VIN mapping, improving the overall quality and safety of vehicles.
• Timely Delivery: Vehicles can be delivered to customers on schedule, enhancing customer satisfaction and trust.
• Regulatory Compliance: The system ensures that vehicles adhere to regulatory standards, reducing the risk of compliance issues or recalls.

The VIN plate mapping system during cockpit assembly and quality checks represents a pivotal advancement in automotive manufacturing. It embodies the commitment to precision, quality, and compliance that are foundational to the industry. By seamlessly integrating Scanflow with the manufacturing process, manufacturers can enhance their operational excellence, meet customer expectations, and uphold the highest standards of safety and quality in the vehicles they produce.