TECHNOLOGY

NIE Neurala Intelligence Engine

The patent-pending, bio-inspired Neurala Intelligence Engine (NIE) is what makes devices smarter. It enables deep autonomy on low-powered, industry-standard processors.  It also enables real-time learning and tracking, object classification and autonomous operation and adaptaion. 

Unlike other methods, the Neurala Intelligence Engine efficiently uses low cost “passive” sensors the way an animal does (a camera like eyes, microphones as ears, etc.). 

NIE is part of a full-stack solution offered by Neurala, including a graphical user interface and device control.

Neurala’s NIE is based on the research done by Neurala for NASA and the U.S. Air Force in conjunction with the Boston University Neuromorphics Lab, which studies the way the brain works and embeds the derived fundamental principles in bio-inspired computers and robots.

Neurala holds two fundamental patents (US 8,648,867 and 9,189,828) that covers hardware and software computational models, often called Artificial Neural Networks (ANN), running on graphic processing units (GPU).  The invention is seen as an important foundation for real-time artificial intelligence and robotics applications.

Neurala’s Technology Makes Machines More Useful

ok

REAL-TIME
LEARNING

Does not require advanced training on thousands of images

ok

PASSIVE OBSERVATION

Uses cameras and microphones instead of expensive sensors

ok

LOCAL COMPUTATION

Works without the need for wi-fi or 4G connections

ok

LOWER HARDWARE COSTS

Uses less power for smaller batteries and lighter weight 

Neurala's tech vs. alternatives

TECH

Live video/still images, sounds, other sensors (e.g., IMUs, LIDAR when available.

NEURALA

Neurala can take a data feed directly from the sensors on your smart machine.

OTHERS

Mostly just still images processed one-by-one on a server. Most solutions pipelined (only 1 capability, perception or navigation or audition)


Local intelligence


Neurala can process the video using the processor found in the device of controller, saving power and cost.


Server processing.


Local communications


Since Neurala processes locally, there is no communications delay or risk of lost connection.


Wifi or 4G communications to host


What and where


Neurala can tell you what something is and where it is located in 3D space so that your smart machine or user can interact with it.


What only, not where


Learning in real-time


Neurala’s real-time learning means that smart machines can learn the objects that matter in their own environment on-the-fly (in seconds).


Batch learning, time consuming, uses servers, expensive and connectivity-dependent


Incremental learning


Neurala can keep learning every time it sees the object in order to get even more accurate.


Re-learning each time


Integrated functionality


Vision, audio, navigation, motor control, work together


Each sense operates independently


Programmability and customization


UI & customization of machine behaviors


No UI or programmability layer

Neurala's tech vs. alternatives

TECH

NEURALAOTHER

Live video/still images, sounds, other sensors (e.g., IMUs, LIDAR when available.

Neurala can take a data feed directly from the sensors on your smart machine.

Mostly just still images processed one-by-one on a server. Most solutions pipelined (only 1 capability, perception or navigation or audition)

Local intelligence

Neurala can process the video using the processor found in the device of controller, saving power and cost.

Server processing.

Local communications

Since Neurala processes locally, there is no communications delay or risk of lost connection.

Wifi or 4G communications to host

What and where

Neurala can tell you what something is and where it is located in 3D space so that your smart machine or user can interact with it.

What only, not where

Learning in real-time

Neurala’s real-time learning means that smart machines can learn the objects that matter in their own environment on-the-fly (in seconds).

Batch learning, time consuming, uses servers, expensive and connectivity-dependent

Incremental learning

Neurala can keep learning every time it sees the object in order to get even more accurate.

Re-learning each time

Integrated functionality

Vision, audio, navigation, motor control, work together

Each sense operates independently

Programmability and customization

UI & customization of machine behaviors

No UI or programmability layer

×

Selfie Dronie Feature list

FLIGHT FEATURES

NEURALA INTELLIGENCE FEATURES

AUTOMATIC FEATURES

VIDEO AND SHARING FEATURES

SAFETY FEATURES

WARNINGS

DISPLAY

ALWAYS AVAILABLE HELP SYSTEM

MULTI-LINGUAL CONTROLS AND USER CONTROLS

×