Newton by Archetype AI

A first-of-its-kind AI model that understands the physical world.

Contact for Pricing

Description

Newton, developed by Archetype AI, is a pioneering foundation model designed to perceive and reason about the physical world in real-time. Referred to as a Large Behavior Model (LBM), it operates by fusing data from multiple sensor types—such as radars, cameras, accelerometers, and gyroscopes—with natural language understanding. This approach allows Newton to identify and interpret complex patterns and behaviors in the physical environment that are often too fast or subtle for human perception.

The core of Newton lies in its ability to encode diverse sensor signals into a universal embedding space, creating a compressed, real-time mathematical representation of the physical world. This encoded information can then be decoded into various formats like text summaries, visualizations, or code, providing actionable insights. Users can interact with Newton using natural language, speech, or gestures to query sensor data, receive predictions, or automate tasks. It offers an API for integration, supports private data handling and fine-tuning, and allows for the creation of specialized edge models for deployment on devices or machinery.

Key Features

Physical World Understanding: Fuses multimodal sensor data (radars, cameras, time-series sensors) and natural language.
Large Behavior Model (LBM): Proprietary foundation model perceives and reasons about physical behavior.
Universal Embedding Space: Encodes diverse sensor signals into a single real-time representation.
Generative Output: Decodes sensor meaning into text, visualizations, or code.
Multimodal Interaction: Supports interaction via natural language, speech, or gestures.
API Access: Provides an API to integrate physical AI capabilities into applications.
Sensor Agnosticism: Compatible with various sensors including cameras, microphones, radars, and time-series sensors.
Data Privacy & Ownership: Allows running the model with proprietary data on user or Archetype infrastructure, ensuring data privacy.
Private Fine-Tuning: Enables customization with proprietary data for specialized use cases.
Edge Model Creation: Supports generating smaller models for deployment on edge devices.

Use Cases

Improve safety monitoring on construction sites.
Enhance smart home automation based on presence, focus, and preferences.
Develop adaptive vehicle interactions responding to occupants and gestures.
Optimize factory operations through real-time anomaly detection and predictive maintenance.
Monitor energy infrastructure for predictive maintenance and optimized production.
Generate comprehensive healthcare reports from wearable and home device data.