Announcement: PlusAI and Texas Ventures Acquisition III Corp Announce Filing of Registration Statement on Form S-4 Read more →
Contact Us
AV 2.0

Our Software Architecture

Three-layer redundancy architecture delivers robust physical AI for autonomy.

AV2.0 Three-Layer Architecture

PrimaryDriving SystemRedundantFallback SystemRemoteOperation SystemCommandsVehicleNavigationSensorsReasoningVision Language ModelHigh-Level Guidance and DecisionReflexEnd-to-End Transformer ModelPerceptionMotion ForecastScene AbstractionGuardrailsSafety Assurance ModelVerified Planningand ControlRedundancyFail-Safe Fallback SystemRemote OperationHuman-in-the-Loop Monitoring and Support
Edge
Primary Driving System
Commands→ Reasoning
Vehicle→ Reflex
Navigation→ Reflex
Sensors→ Guardrails
Inputs
Reasoning
Vision Language Model
High-Level Guidance and Decision
Reflex
End-to-End Transformer Model
PerceptionMotion Forecast
Scene Abstraction
Guardrails
Safety Assurance Model
Reflex + Guardrails
Verified Planning and Control
Redundant Fallback System
Redundancy
Fail-Safe Fallback System
Cloud
Remote Operation System
Remote Operation
Human-in-the-Loop Monitoring and Support

SuperDrive™’s three-layer architecture is designed to provide redundancy and fail-safe behavior. It runs paid freight operations with a safety driver in Texas today.

The REASONING–REFLEX Dual-Model

Layer 1: Primary Edge Computer

REASONING

Vision Language Model in action

Leverages a Vision Language Model to interpret complex real-world interactions and generate high-level driving decisions for out-of-ODD edge cases. Acts as strategic driving intelligence, understanding context and making informed decisions for unusual scenarios.

REFLEX

End-to-end model in action

Our transformer-based end-to-end model fuses perception with motion planning to execute rapid, real-time maneuvers. Our network design allows us to provide high-performance models that can rapidly scale up to automotive grade compute.

Multi-Redundancy

Line drawing of a PlusAI truck showing where its sensors are: the primary driving system (cameras, lidar and radar, with radar shared from the fallback system) in teal, and the redundant fallback system (cameras and radar) in green.
Layer 2: Secondary Edge Computer

Redundant Fallback System

Operates on a secondary edge computer that continuously monitors the primary system’s health. If a fault is detected, it is designed to trigger minimal-risk fallback maneuvers to bring the vehicle safely to a stop.

Remote operations view of three PlusAI trucks, each showing live surround-camera feeds, an in-cab camera and the SuperDrive dashboard with current speed and speed limit.
Layer 3: Cloud

Remote Operations

Provides mission control and human-in-the-loop support for scenarios beyond the virtual driver’s Operational Design Domain (ODD), ensuring oversight and intervention when needed. Remote operation activity will be handled by our OEM partners and not by PlusAI directly.

Key Advantages

  • Interpretability

    Through semantic reasoning via a Vision Language Model and a defined perception layer, SuperDrive™’s architecture is designed to make driving decisions transparent and explainable for engineers and stakeholders.

  • Traceability

    The architecture enables full traceability, making it possible to trace and rectify errors, identify corner cases, and support continual improvement across the entire autonomy stack.

  • Enhanced Safety

    Rule-based safety envelopes enable region-specific driving behavior while cloud-based remote operations provide a human-in-the-loop for scenarios beyond the virtual driver’s ODD.

  • Scalability

    Zero-shot learning capabilities allow SuperDrive™ to adapt to out-of-sample edge cases, diverse regions, and new vehicle platforms without extensive retraining.

Tech Insights

Read more

Let’s discuss how AV 2.0 can fit into your autonomy plans.

Contact Us