Is this page helpful?

NVIDIA Halos Outside-In Safety Blueprint#

Introduction#

What is Outside-In Safety?#

Onboard sensors have a limited field of view when dealing with occlusions like static walls, payloads, and moving objects such as workers and vehicles. With an outside-in safety agent, robot awareness expands beyond the robot’s eyes by communicating with sensors and cameras placed throughout the facility.

This broader 360° view helps agents understand, predict, and monitor incredibly complex environments that onboard sensors might miss. The result is more proactive protection, fewer slowdowns, and safer, more efficient operations.

Overview#

NVIDIA Halos Outside-In Safety Blueprint extends robot perception beyond onboard sensors by using external infrastructure cameras, AI perception, and safety logic to deliver real-time functional safety to maximize operational throughput. Running on NVIDIA IGX, it enables robots to safely operate alongside workers at high efficiency modes while dynamically adapting to complex environments.

  • Monitor safety parameters of an industrial environment in real-time

  • Detect potential safety events proactively

  • Communicate safety events to corresponding mitigation mechanisms

  • Maintain continuous safety verification

  • Provide reliable logging and auditing capabilities

Key Features#

  1. Modular Architecture

    • Built as a modular framework, allowing system integrators to seamlessly integrate it into targeted environments

    • Adaptive to a variety of sensor processing pipelines

  2. Safety-Centric Design

    • Designed adhering to safety standards

  3. Real-time Capabilities

    • Designed to handle real-time safety monitoring and decision-making

    • Supports low-latency communication between components

    • Implements priority-based message handling for critical safety events

  4. Scalable Implementation

    • Can be implemented in small embedded systems to large distributed systems

    • Supports various communication protocols and data formats

    • Allows for incremental safety feature implementation

This documentation covers:

  • Getting Started: Bring up a first working system, from the AI perception pipeline to the Safety Core.

  • Deployment Guide: Install, run, configure, and troubleshoot the Safety Core in target environments.

  • Integration Guide: Understand the detailed system design, components, and data flows, and build custom applications.

  • Closed-Loop Testing: Configure and execute closed-loop testing of the system, in simulation on a single host or with the Safety Core on target hardware.

  • Reference: Look up launcher options, configuration keys, runtime profiles, and terminology.

API Documentation#