Why We Use AI Visual Navigation?

Why We Use AI Visual Navigation?

TerraVision: The Latest Technology Leading the 3rd-Gen Robotic Lawn Mower Revolution

Perimeter wires, RTK signals, and combinations of RTK with visual systems are common navigation methods used in today’s robot lawnmowers. However, we take a different approach with the sole AI vision system called “TerraVision”. What is “TerraVision” and what makes it stand out? Learn more now below!

 

Why Choose AI Visual Navigation?

A robotic lawn mower's navigation system is crucial, influencing both the ease of setup and mowing efficiency, ultimately shaping the user experience. We believe visual navigation represents the future of robotic lawnmowers.

 

The following is a quick comparison of how different navigation systems affect setup and mowing efficiency:

Perimeter Wires RTK Signals

TerraVision (AI Vision)

 Setup

1. Base station

2. Perimeter wires

1. Base station

2. Perimeter wires

Base station

 Mapping

No mapping

Manual mapping

Autonomous AI mapping

 Mowing Route

Chaotic mowing path Organized mowing path Organized mowing path
 Self Positioning Without self-positioning May have issues in narrow places or under covers

Precise self-positioning

 Night Mowing Capable Capable Not recommended

 

What Is TerraVison?

TerraVision is a proprietary, all-in-one vision system for autonomous AI mapping, real-time 3D obstacle avoidance, and precise self-positioning. It is a dedicated system that combines both top-grade cameras and cutting-edge technologies, elevating your lawn care experience.

 

Automotive-Grade Camera System

TerraMow features a ray of automotive-grade cameras designed to capture its surroundings and necessary supporters coping with the extreme weather. Therefore, it excels in diverse working conditions, providing unmatched visibility and adaptability for seamless operation.

 

(1) Main cameras

Featuring three high-resolution automotive-grade cameras to preserve surroundings:

- 2 Stereo 3D cameras for detecting the accurate distance between the TerraMow and surrounding obstacles;

- A semantic AI camera for capturing details of the surrounding environment in pixel-level details. 

 

(2) Supporters

- An LED light activates automatically for low-light conditions;

- An adaptive defogger can evaporate water on the TerraVision panel after dew or rain.

 

Advanced Visual Perception Technologies

TerraVision utilizes proprietary 3D visual sensing and AI semantic perception to surpass conventional 3D vision systems. This technology enables TerraMow to perceive and classify 3D objects in real time, transforming how it navigates and interacts with its surroundings for smarter, more efficient lawn care.

 

(1) The most powerful 3D perception technology

TerraVision surpasses the capabilities of traditional 3D lidar systems by perceiving 3D points with a 98X higher density and by providing 400X more information density. Therefore, TerraMow gains a precise and accurate understanding of nearby tiny obstacles (as mini as 10*10*5cm). Not only capturing and creating detailed 3D models but also accurately measuring the exact distance between them.

 

(2) Pixel-perfect AI semantic perception technology

TerraMow goes beyond perceiving obstacles in 3D models by providing a clear real-world visual representation and accurate identification of objects with the AI semantic perception technology. With its high-resolution imaging capabilities of up to 1920x1080, TerraMow perceives the surrounding environment as well as analyzes it using its AI Neural Network. TerraMow can generate a lawn map with pixel-level precision, facilitate precise navigation, and enhance obstacle recognition for optimal performance.

 

(3) Hybrid fusion vision

TerraMow perceives its surroundings both with the 3D perception technology for learning the exact shape and distance of the surroundings and with the AI semantic perception technology for capturing real-world pictures and identifying what they are.

 

Combining the information perceived by these two technologies, TerraMow can recognize where the lawn boundaries and obstacles are, and update the information every time it is on mowing tasks. That’s why TerraMow can explore your lawn autonomously and efficiently, yet with rich details all on its own.

 

(4) Biomimetic neural vision positioning system

TerraMow utilizes an innovative biomimetic neural vision positioning system to enable precise and real-time self-positioning. Functions exactly like human navigation, TerraMow continuously identifies and extracts landmarks (a leaf, a flower, or a ground light fixture) from the images for self-positioning, allowing the system to learn and adapt to the dynamic changes of your lawn.

 

Therefore, you no longer have to use additional positioning devices or markers, such as RTK stations, to help the robot position itself. Furthermore, by using various landmarks and references on your lawn, TerraMow seldom gets lost due to the annoying RTK(GPS) signal issues, allowing it to cover those challenging landscapes with ease., such as, under thick tree covers, roofs, and narrow pathways.

 

Complex Garden Mastery

Powered by TerraVision, TerraMow is the world’s first truly hassle-free intelligent robotic lawn mower, making it one of the top choices for complex gardens thanks to:

  • Precise autonomous lawn mapping and exploration
  • Expert navigation around complex obstacles
  • Accurate positioning unaffected by environmental factors
  • Systematic cutting patterns for consistent and efficient mowing

 

Reading next

Leave a comment

This site is protected by hCaptcha and the hCaptcha Privacy Policy and Terms of Service apply.