{"id":8,"date":"2023-09-17T12:49:10","date_gmt":"2023-09-17T12:49:10","guid":{"rendered":"https:\/\/renewpole.com\/?page_id=8"},"modified":"2026-06-04T13:59:38","modified_gmt":"2026-06-04T18:59:38","slug":"home","status":"publish","type":"page","link":"https:\/\/avexit.com\/","title":{"rendered":"Home"},"content":{"rendered":"\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

SWAP BY AI\u2122<\/h1>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

One platform, many lives.<\/h1>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

Transformation Behavior for Physical AI<\/h1>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

The future of Physical AI for mobility will be transformative: modular vehicles that can autonomously reconfigure themselves. Anywhere.<\/em><\/p>

At AV Exit, we\u2019re not building vehicles. We invented the transformation behavior IP. <\/strong>Swap by AI\u2122 defines a new approach to modular vehicles for the autonomous age, conceptually based upon a longer-life Power Platform<\/strong> (chassis, battery, motors, and wheels engineered for 20+ years of service, plus AI compute and cameras to drive autonomously when the pod is detached) and a separable Passenger Pod<\/strong> (the customizable cabin, likely refreshed every 3 to 4 years to keep pace with advancing AV compute and evolving consumer needs).<\/p>

The Swap by AI\u2122 Family<\/strong><\/p>

Swap by AI\u2122\u00a0 \u2022\u00a0 Enter by AI\u2122\u00a0 \u2022\u00a0 Safe by AI\u2122\u00a0 \u2022\u00a0 Charge by AI\u2122\u00a0 \u2022\u00a0 Exit by AI\u2122<\/p>

Each behavior is described in The Swap by AI\u2122 Advantage section below. Patent status detail in the Patent Portfolio summary.<\/p>

Swap by AI\u2122 in <\/strong>NVIDIA\u2019s ecosystem: AlpaSim can test it. OmniDreams can render it. AlpaGym can train it. Alpamayo can reason about it. DRIVE AGX Thor can embody it. DGX can refine it through post-deployment fleet data. The future is now.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t

\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t<\/path>\t\t<\/path>\t\t<\/path>\t<\/g><\/g><\/svg>\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

Physical AI in Action<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t\t
\n\t\t
\n\t\t\t\t\t
\n\t\t
\n\t\t\t\t\t
\n\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

Physical AI<\/strong> is artificial intelligence that doesn\u2019t just describe the physical world but acts on it: perceiving conditions, reasoning about response options, and executing physical change. NVIDIA’s broader stack (Omniverse and Cosmos 3 for simulation, DRIVE AGX Thor for embodiment, DGX for refinement) has made the compute foundation possible \u2014 and the Alpamayo Open Portfolio (the Alpamayo model, the AlpaSim simulation framework, the AlpaGym training framework, and the Physical AI Dataset), with the Cosmos-based OmniDreams world model, now turns that compute into AV-specific tooling. The behaviors that vehicles will perform under that compute remain to be defined.<\/p>

AV Exit\u2019s IP defines transformation behavior.<\/strong> Not as software running on top of NVIDIA\u2019s stack, but as patented and patent-pending architectural definitions for a new approach to modular autonomous vehicles.<\/p>

FIVE SCENES FOR FIVE BEHAVIORS<\/strong><\/p>

Picture an autonomous vehicle pulling into a driveway, detaching its cabin onto temporary supports, accepting a different cabin configuration, and driving away. Without human intervention. Without specialized infrastructure. On a flat surface that didn\u2019t need to be prepared for the operation. That\u2019s Swap by AI\u2122.<\/strong><\/p>

Picture a vehicle whose passenger doors open and close not via dedicated motors but via the same wheels that drive it down the road, eliminating an entire mechanical subsystem and its lifetime failure modes. That\u2019s Enter by AI\u2122.<\/strong><\/p>

Picture a vehicle that perceives an imminent, unavoidable impact, and partially separates its cabin from its chassis to attenuate the collision forces transmitted to occupants. That\u2019s Safe by AI\u2122.<\/strong><\/p>

Picture a parked vehicle with a solar-energy collecting photovoltaic roof, whose cabin detaches and tilts onto support legs, revealing additional photovoltaic surface on the platform itself, while the cabin remains usable for working, sleeping, or storage. That\u2019s Charge by AI\u2122.<\/strong><\/p>

Picture a vehicle whose AI detects an unavoidable impact and, instead of bracing for it, ejects the entire passenger pod along a pre-computed safer trajectory. That\u2019s Exit by AI\u2122.<\/strong><\/p>

THE ALPAMAYO OPEN PORTFOLIO<\/strong><\/p>

NVIDIA’s Alpamayo,<\/strong> the autonomous-vehicle reasoning model announced by Jensen Huang at CES 2026 in January, is the first NVIDIA full-stack AV deployment to reach a production vehicle: the all-new Mercedes-Benz CLA, debuting NVIDIA DRIVE AV with a U.S. Level 2++ launch in 2026.<\/p>

At GTC 2026 Taipei in June, the model family advanced to Alpamayo 2 Super (an open 32-billion-parameter reasoning model with full 360-degree perception). Alpamayo demonstrates that AVs can be trained to reason from a few million simulation miles rather than billions of real-world miles.<\/p>

NVIDIA’s AlpaSim,<\/strong> the open-source closed-loop simulation framework released at CES 2026 and available on GitHub, orchestrates the environments in which Alpamayo’s policy decisions get tested in real time.<\/p>

NVIDIA’s OmniDreams,\u00a0<\/strong>real-time generative closed-loop AV simulation built on Cosmos, which NVIDIA previewed as “AlpaDreams” at GTC 2026 San Jose in March and presented as OmniDreams at GTC 2026 Taipei in June, generates the photorealistic, action-conditioned views inside the simulation loop. Text-prompt scenario authoring lets developers describe what should happen and then watch the simulated “ego” AV respond.<\/p>

NVIDIA’s AlpaGym,<\/strong> the closed-loop reinforcement-learning framework introduced at GTC 2026 Taipei in June, post-trains the policy through continuous decision-and-observation cycles inside AlpaSim.<\/p>

NVIDIA\u2019s Physical AI Dataset,<\/strong> another component of the Alpamayo open portfolio, supplies the real-world driving data Alpamayo learns from: 300,000 clips spanning 1,700 hours across 2,500 cities in 25 countries. The Dataset captures what fleets do today (lane following, merging, braking, urban navigation). Transformation behaviors lie beyond what any fleet performs yet.<\/p>

AV Exit’s IP is the missing content for this open portfolio.<\/strong> These behaviors cannot be learned from observation, but they can be defined architecturally, run as closed-loop scenarios in AlpaSim, rendered photorealistically by OmniDreams, post-trained through AlpaGym, and used to train Alpamayo’s reasoning. AV Exit’s patents define the architecture. AlpaSim runs the closed loop. OmniDreams renders the scenes. AlpaGym post-trains the policy. Alpamayo learns the behaviors. DRIVE AGX Thor executes them on modular vehicle hardware.<\/span><\/p>

THE STRATEGIC MOMENT<\/strong><\/p>

The EV transition has been disrupted by policy shifts, tariff regimes, charging infrastructure gaps, raw materials volatility, demand shortfalls, and write-downs across multiple legacy automakers. The industry pattern is now visible: incremental EV strategies face structural headwinds that no single OEM can solve on its own timeline. The strategic question for an OEM facing that pattern is not how to fix EV-1.0. It\u2019s how to leapfrog past it into AV Level 4 (hands-off, eyes off), and beyond.<\/p>

Physical AI is the leapfrog.<\/strong> It is built to move<\/em>. Modular autonomous vehicles built on NVIDIA\u2019s compute stack and AV Exit\u2019s transformation behavior IP represent a new paradigm: vehicles whose architectural moves create capabilities no unibody vehicle can mechanically perform. An OEM committing to this approach isn\u2019t catching up on EVs. It\u2019s defining the next category while competitors are still solving the last one. OEMs deploying self-reconfiguring vehicles will have a competitive advantage.<\/p>

VALIDATION PATHWAY<\/strong><\/p>

NVIDIA’s DRIVE Hyperion has become the common platform for the autonomous age. At GTC Taipei in June 2026, Jensen Huang said car manufacturers representing roughly 80% of the world’s vehicles have adopted DRIVE Hyperion, NVIDIA’s Level 4-ready platform \u2014 the OEMs that would build modular AVs, and the manufacturing base into which Swap by AI deploys. BYD, Geely, Nissan, and Hyundai (four of the world’s top ten automakers) have adopted DRIVE Hyperion for Level 4 development. This is the ecosystem AV Exit’s transformation behavior IP is built for, and it is being deployed at global scale.<\/p>

The NVIDIA-Hyundai partnership announced in October 2025 anchors South Korea’s National Physical AI Cluster, a public-private partnership deploying roughly $3 billion and 50,000 NVIDIA Blackwell GPUs. The partnership encompasses Kia, whose Platform Beyond Vehicle line ships modular platforms with swappable pods. Kia’s planned Easy Swap system requires dedicated conversion centers. Swap by AI\u2122 enables autonomous coupling anywhere, no depot required. Enhanced consumer convenience. Reduced OEM infrastructure cost.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t

\n\t\t\t\t\t
\n\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

CONCEPT GALLERY<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

These concept renderings illustrate our patented and patent-pending technologies. The Swap by AI\u2122 family of transformation behaviors can be integrated with most modular designs.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t

\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t\t\t
\"Swap
Swap by AI\u2122 - the key to self-reconfiguring modular vehicles<\/figcaption><\/figure><\/div>
\"Power
Power Platform: Multi-million-mile autonomous chassis with Charge by AI\u2122 solar collection surfacing<\/figcaption><\/figure><\/div>
\"Swap
Swap by AI\u2122 Precision - AI-controlled alignment<\/figcaption><\/figure><\/div>
\"Independent
Independent Pod Operation via Swap by AI\u2122 retractable stands<\/figcaption><\/figure><\/div>
\"Swap
Swap by AI\u2122 Versatility - Same platform, different pods<\/figcaption><\/figure><\/div>
\"Safe
Safe by AI\u2122 - Electromagnetic impact absorption<\/figcaption><\/figure><\/div>
\"Exit
Exit by AI\u2122 - Emergency safety deployment aftermath<\/figcaption><\/figure><\/div>
\"Swap
Swap by AI\u2122 - the dawning of a new mobility paradigm<\/figcaption><\/figure><\/div>\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t<\/i>\t\t\t\t\t<\/div>\n\t\t\t\t\t
\n\t\t\t\t\t\t<\/i>\t\t\t\t\t<\/div>\n\t\t\t\t\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

THE SWAP BY AI\u2122 ADVANTAGE<\/h1>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

Four virtues, fulfilled by five branded behaviors.<\/h1>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

VERSATILITY\u00a0<\/strong><\/p>

A single Power Platform serves fundamentally different use cases across its life: passenger, cargo, work, family, recreation. Each via a different pod. Swap by AI\u2122 enables autonomous pod exchange anywhere a flat surface exists. No dedicated swapping station required. As Swap by AI\u2122 inventor Dave Carroll puts it: “mobile office for the week, pickup pod for the weekend.” Or as our slogan puts it: One platform, many lives.<\/em><\/p>

SAFETY<\/strong><\/p>

Modular AVs gain crash-response options that unibody vehicles mechanically cannot perform. Safe by AI\u2122 executes tactical partial separation under collision conditions, with optional electromagnetic impact absorption attenuating remaining forces. Exit by AI\u2122 deploys the pod away from imminent collision along an AI-coordinated safer path. Both are deliberate vehicle-initiated responses, extending the vehicle\u2019s response toolkit beyond what crumple zones and steering avoidance alone can offer.<\/p>

VALUE<\/strong><\/p>

A Power Platform engineered for 20+ years amortizes the costliest components across multiple cabin generations. This may allow for longer-term platform financing and thus lower monthly payments for the vehicle overall. OEMs earn recurring revenue from pod refresh cycles every 3 to 4 years rather than from one-time vehicle sale. Consumers keep the platform for decades and refresh the cabin to keep pace with advancing AV compute. NVIDIA\u2019s silicon presence scales accordingly: Thor count per vehicle increases from 2 (unibody) to up to 4 (Power Platform plus Pod), with additional Thors arriving as pods refresh across the platform\u2019s lifecycle. Enter by AI\u2122 uses the platform\u2019s own wheels to actuate pod doors, removing dedicated door motors and the associated parts count.<\/p>

SUSTAINABILITY<\/strong><\/p>

Charge by AI\u2122 reveals the Power Platform\u2019s photovoltaic surface when the pod separates, with AI coordinating timing, positioning, suspension tilt, and energy transfer. By exposing surface area normally covered when assembled, this transformation behavior effectively multiplies the vehicle\u2019s solar collection capacity. The architecture reduces lifetime resource consumption: one durable platform serves multiple cabin generations rather than each generation requiring a complete new vehicle.\u00a0<\/p>

PATENT PORTFOLIO<\/strong><\/p>

Five branded behaviors anchor a 13-application portfolio.<\/p>

Swap by AI\u2122 –<\/strong>\u00a0U.S. Patent 12,515,705 (issued January 2026)<\/p>

Enter by AI\u2122<\/strong> – U.S. Patent 12,637,028 (issuing May 26, 2026)<\/p>

Exit by AI\u2122<\/strong> – U.S. Patent 11,618,474 (issued April 2023)<\/p>

Safe by AI\u2122<\/strong> – Provisional patent application filed April 2026<\/p>

Charge by AI\u2122<\/strong> – Provisional patent application filed April 2026<\/p>

Additional applications cover specific implementation methods. International filings active in PCT, China, Europe, and South Korea.<\/em><\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t

\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t<\/path>\t\t<\/path>\t\t<\/path>\t<\/g><\/g><\/svg>\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

THE TEAM BEHIND SWAP BY AI\u2122 <\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

Founded in 2022 in Grantsburg, Wisconsin, AV Exit, LLC<\/strong> owns the Swap by AI\u2122 family of patented and patent-pending transformation behaviors.\u00a0<\/p>

DAVE CARROLL, FOUNDER\/CEO<\/strong><\/p>

Co-founded RAM Center (FANUC\u2019s multi-year top distributor) in 1986. Forty years of patented invention across industrial and retail robotics, wearable computing, LED lighting, and advanced manufacturing education. Sole inventor of the Swap by AI\u2122 family.<\/p>

IAN D. CARROLL, PRESIDENT<\/strong><\/p>

Leads strategy, partnerships, and acquisition positioning. Background in research and strategy for crisis communications. Engaging NVIDIA, OEM ecosystem partners, and strategic investors for acquisition and partnership discussions.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t

\n\t\t\t\t\t
\n\t\t
\n\t\t\t\t\t
\n\t\t
\n\t\t\t\t\t
\n\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t
\n\t\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

SWAP BY AI\u2122 – <\/strong>Transformation Behaviors for Physical AI<\/p>

For Strategic Investment or Acquisition, Contact:<\/em><\/p>

Dave Carroll, Founder\/CEO<\/strong>
AV Exit, LLC<\/p>

507.261.8937<\/div>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t\n\t\t\t<\/i>\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t\n\t\t\t<\/i>\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t\n\t\t\t<\/i>\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t\t
\n\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

Swap by AI\u2122, Enter by AI\u2122, Safe by AI\u2122, Charge by AI\u2122, and Exit by AI\u2122 are trademarks of AV Exit, LLC. Protected by U.S. Patents 12,515,705 and 11,618,474, plus U.S. Patent 12,637,028 issuing May 26, 2026, with additional applications pending and international filings active. \u00a9 2026 AV Exit, LLC. All Rights Reserved.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t

\n\t\t\t\t\t
\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"

SWAP BY AI\u2122 One platform, many lives. Transformation Behavior for Physical AI The future of Physical AI for mobility will be transformative: modular vehicles that can autonomously reconfigure themselves. Anywhere. At AV Exit, we\u2019re not building vehicles. We invented the transformation behavior IP. Swap by AI\u2122 defines a new approach to modular vehicles for the… Continue reading Home<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":1082,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"elementor_header_footer","meta":{"footnotes":""},"class_list":["post-8","page","type-page","status-publish","has-post-thumbnail","hentry","entry"],"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/avexit.com\/index.php?rest_route=\/wp\/v2\/pages\/8","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/avexit.com\/index.php?rest_route=\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/avexit.com\/index.php?rest_route=\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/avexit.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/avexit.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=8"}],"version-history":[{"count":117,"href":"https:\/\/avexit.com\/index.php?rest_route=\/wp\/v2\/pages\/8\/revisions"}],"predecessor-version":[{"id":1389,"href":"https:\/\/avexit.com\/index.php?rest_route=\/wp\/v2\/pages\/8\/revisions\/1389"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/avexit.com\/index.php?rest_route=\/wp\/v2\/media\/1082"}],"wp:attachment":[{"href":"https:\/\/avexit.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=8"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}