Vehicles that transform themselves. Anywhere.
Swap by AI™ (patent issued January 2026) enables modular autonomous vehicles to physically reconfigure themselves by using only their own motors, wheels, and AI decision-making to swap entire passenger and cargo pods. Any flat surface. Many possible pod configurations. No dedicated swapping stations required.
The Swap by AI™ IP Family
Enter by AI™ – wheel-actuated pod door operation (allowed February 2026)
Safe by AI™ – tactical partial separation with optional electromagnetic impact absorption (provisional filed April 2026)
Charge by AI™ – solar collection via autonomous platform exposure (provisional filed April 2026)
Exit by AI™ – (our first issued patent, April 2023) deploys pods away from imminent collisions proactively along a safer path
Platform Model Offering: We’re not building vehicles. We’re defining the transformation behavior IP that lets modular autonomous vehicles reconfigure themselves.
Swap by AI™ represents a new Physical AI paradigm for NVIDIA’s autonomous-vehicle stack: Omniverse with Cosmos can simulate it, DRIVE AGX Thor can embody it, and DGX can refine it through post-deployment fleet data.
Physical AI — machines that autonomously manipulate the real world — is projected to create $4 trillion in value by 2030 (McKinsey). In 2025, NVIDIA CEO Jensen Huang predicted autonomous vehicles would be “the first multi-trillion-dollar robotics industry.” NVIDIA’s Alpamayo, unveiled at CES 2026, is the first reasoning AI for autonomous vehicles, built to decompose novel scenarios into familiar components.
AV Exit’s transformation behaviors are deliberate capabilities the vehicle initiates rather than edge cases it navigates, and the simulation scenarios needed to develop them don’t yet exist. No vehicle has ever performed emergency pod separation at highway speed, autonomous coupling in unstructured environments, or coordinated solar collection while the pod is detached. These transformation behaviors are simulation-native by necessity.
Once deployed, fleet data starts flowing back, intensifying training demand across NVIDIA’s Data Center segment. Transformation behaviors get refined against real-world ground truth no synthetic dataset could fully predict. Until then, NVIDIA’s Cosmos pipeline, now including AlpaDreams for closed-loop scenario authoring, is the only way to get these behaviors road-ready. NVIDIA Halos provides the safety certification path before they leave the simulator.
Acquiring AV Exit’s IP gives NVIDIA the missing piece to bring to its DRIVE ecosystem partners: the transformation behavior IP that could make modular autonomy viable at scale.
One Possible Validation Path, Among Many: Anchored by the NVIDIA-Hyundai partnership announced October 2025, Korea’s National Physical AI Cluster creates a validation pathway for AV Exit’s approach to Physical AI. Hyundai Motor Group and South Korea’s government are co-investing approximately $3 billion in facilities and computing infrastructure (including 50,000 NVIDIA GPUs); NVIDIA also contributes its DRIVE platform and expertise.
The NVIDIA-Hyundai partnership already encompasses Kia, whose PBV line features modular platforms with swappable pods. As conceived today, Kia’s “Easy Swap” system would require dedicated conversion centers for pod exchange. As Kia PBV goes autonomous, infrastructure costs would scale into the billions. Swap by AI™ enables autonomous coupling anywhere, no depot required, yielding both enhanced user convenience and reduced OEM cost. This could then become the model for others, as there are many additional OEM and Tier-1 partners within NVIDIA’s vast automotive ecosystem who could develop products based in part on AV Exit’s IP under the NVIDIA umbrella.
Swap by AI™ (the foundation)
Autonomous pod swapping using the vehicle’s own systems:
• AI analyzes environment and calculates approach
• Motors and wheels execute precise alignment
• Automated connection and disconnection sequence
• Complete verification through sensor fusion
• US Patent 12,515,705 (issued January 2026)
Works anywhere a reasonably flat surface exists: driveways, parking lots, roadsides, depot floors. Any flat surface becomes a reconfiguration point. No dedicated swapping stations required.
Enter by AI™ (wheel-actuated pod door operation)
Routine pod door operation that uses the vehicle’s own wheels rather than dedicated door motors:
• Wheel motion drives pod door open and closed
• Mechanical efficiency reduces parts count and failure model
• Integrates with occupant monitoring for safe operation
• Patent allowed February 2026
Safe by AI™ (tactical partial separation with EMA)
For collision scenarios where full pod deployment isn’t possible or necessary:
• Partial pod-platform separation creates a collision buffer zone
• Electromagnetic impact absorption (EMA) attenuates remaining impact forces
• Differential response based on collision geometry and occupant position
• Provisional patent application filed April 2026
Charge by AI™ (autonomous solar collection)
Energy harvesting while the pod is detached:
• Pod separates from platform during charging windows
• Photovoltaic surface on platform exposed to direct sunlight
• AI coordinates timing, positioning, suspension tilt, and energy transfer
• Provisional patent application filed April 2026
Exit by AI™ (advanced safety capability)
The dramatic capability for unavoidable collisions:
• Collision prediction in milliseconds
• Autonomous pod separation (full or partial) away from imminent impact
• Physics-based occupant protection during separation and descent
• US Patent 11,618,474 (issued April 2023)
Patent Protection
Five branded behaviors anchor a 13-application portfolio:
• Swap by AI™: US Patent 12,515,705 (issued January 2026)
• Enter by AI™: Patent allowed February 2026
• Exit by AI™: US Patent 11,618,474 (issued April 2023)
• Safe by AI™: Provisional patent application filed April 2026
• Charge by AI™: Provisional patent application filed April 2026
Additional applications cover specific implementation methods. International filings active in PCT, China, Europe, and South Korea.
These concept renderings illustrate our patented technologies. The Swap by AI™ family of transformation behaviors can be integrated with any manufacturer’s modular designs.
For NVIDIA:
• Massive GPU-accelerated simulation workloads in Omniverse and Cosmos for behaviors with no real-world training data yet
• A new content domain for the AV simulation pipeline: multi-body physics scenarios that extend existing tools into modular vehicle territory
• Per-vehicle Thor count scales from 2 (unibody baseline) to 4 (Power Platform plus Pod), with pod refresh cycles bringing additional Thors across the vehicle’s 20-year platform life
• New pod models and refresh cycles trigger Halos certification work across the vehicle’s 20-year lifecycle, generating recurring AV-safety engagement that unibody AVs require only once
For OEMs:
• Continuous revenue across the platform’s 20-year life through pod refresh cycles, not just one-time vehicle sale
• Higher margins on pods vs. complete vehicles, and faster design cycles on pod-only innovation
• Five distinct behavior patents extend product differentiation: Swap, Enter, Safe, Charge, and Exit each support different OEM strategies
• No depot infrastructure required: Swap by AI works anywhere, unlike conversion-center pod exchange systems like Kia’s Easy Swap
For Consumers:
• Keep your power platform for decades, not years; financial terms could be extended accordingly for a lower monthly payment
• Refresh pod hardware every 3 to 4 years to keep pace with AV compute and feature obsolescence, without bricking the vehicle
• Pod swap as the everyday operating mode: mobile office in the morning, pickup pod over the weekend
• Twice the solar charging while the pod is detached, with onboard AI facilitating enhanced positioning for solar collection
The Bottom Line:
Swap by AI™ and its family of patented behaviors create value at every level: NVIDIA’s Physical AI stack gets a new content domain and per-vehicle compute multiplier, OEMs gain a recurring-revenue product architecture, and consumers get vehicles that stay current, capable, and uniquely safe across their full mechanical lifespan.
AV Exit, LLC owns the Swap by AI™ family of patented transformation behaviors that let modular autonomous vehicles reconfigure themselves: Swap by AI™, Enter by AI™, Safe by AI™, Charge by AI™, and Exit by AI™.
Dave Carroll, Founder/CEO: Founded RAM Center (FANUC’s multi-year top distributor) in 1986. Forty years of patented invention across industrial robotics, wearable computing, LED lighting, and advanced manufacturing education. Sole inventor of the Swap by AI™ transformation behavior family.
Ian D. Carroll, President: Leads strategy, partnerships, and acquisition positioning. Background in crisis communications and business development shapes the company’s approach to deep-tech messaging.
Status: Engaging NVIDIA, OEM ecosystem partners, and strategic investors for acquisition and partnership discussions.
Swap by AI™, Enter by AI™, Safe by AI™, Charge by AI™, and Exit by AI™ are trademarks of AV Exit, LLC. Protected by US Patents 12,515,705 and 11,618,474, with additional applications pending and international filings active. © 2026 AV Exit, LLC. All Rights Reserved.
