F100H takdim ederiz. Eski traktörler için daha geniş destek burada başlıyor. Şimdi Gör >
LiDAR and Vision: The Dynamic Duo Behind Your New Smart Lawn Mower
translate

Whether you're a lawn care enthusiast or just looking to automate a yard chore, there's never been a better time to pick up a robotic lawn mower. With more choices than ever to give your lawn a nice trim, we decided to equip our Sveaverken Blix robotic lawn mower with a combination of LiDAR and vision technology.

In this short overview, we'll outline why our lawn mower is a top pick, and how LiDAR works to intelligently map your lawn.

Intelligence for Ease of Use

The goal was simple: with our new Blix, we wanted to create a lawn mower that requires minimal human intervention. We wanted to bypass the hassle of perimeter wires and satellite-based systems.

To achieve this, we needed a machine that could understand its environment and navigate it intelligently. This is where sensor fusion technology comes in. Information from the LiDAR sensor, an IMU (inertial measurement unit), and a vision sensor enables Blix to understand where it is, where it needs to go, and get the mowing job done.

LiDAR: A Quick Primer

Imagine you're trying to understand a room without being able to see it. You might tap around with a cane to get a sense of where things are. LiDAR is a bit like that, but instead of a cane, it uses laser light.

LiDAR, which stands for Light Detection and Ranging, is a remote sensing method that uses light to measure distances. It works by emitting laser pulses and measuring the time it takes for the reflected light to return to the sensor. This information is used to create a detailed 3D map of the environment.

LiDAR technology has its roots in aerospace and defense applications, where it was used for terrain mapping. Over the years, the technology has become more affordable and accessible, leading to its adoption in various industries, including robotics and autonomous vehicles.

LiDAR Sensors Are Not Created Equal

LiDAR sensors vary in terms of range, resolution, field of view, and cost. This article from Think Autonomous explains the differences between 9 types of LiDAR sensors in great depth, categorizing them by how they scan, the dimensionality of the results produced, and modulation, or how the light waves are sent and received to measure distance.

For the purposes of our discussion, we'll focus on dimensionality, and take a look at the distinction between 2D and 3D LiDAR.

As the names suggest, 2D LiDAR creates a flat representation of the environment, while 3D LiDAR provides a more accurate and detailed 3D map. This difference has implications on the cost and operating environment.

Indoor vs. Outdoor Applications

The choice between 2D and 3D LiDAR is significantly influenced by the environment in which the robot operates.

Indoor environments:

  • Typically, more controlled conditions with less complex terrain.
  • 2D LiDAR can often suffice for tasks like close-range mapping, localization, and obstacle avoidance.
  • However, if the robot needs to interact with objects at different heights (e.g., shelves), 3D LiDAR might be necessary.

This said, 2D LiDAR sensors can be a good choice for:

  • Indoor floor cleaning robots: These robots primarily operate on flat surfaces and need to detect obstacles like walls, furniture, and people. 2D LiDAR is sufficient for creating a map of the floor plan and navigating accordingly.
  • Warehouse and factory automation: In these environments, robots often operate on defined paths and need to avoid static obstacles like shelves and machinery. 2D LiDAR can provide adequate spatial information for these tasks.

Outdoor environments:

  • More dynamic and complex conditions with varying terrain, weather, and lighting.
  • 3D LiDAR is generally preferred due to its ability to handle uneven surfaces, obstacles at different heights (e.g., trees and buildings), and changing lighting conditions.
  • However, in specific outdoor applications with limited vertical obstacles (e.g., lawn mowing on flat terrain), 2D LiDAR could be considered.

For a wireless robotic lawn mower like the Sveaverken Blix, we needed a sensor that could accurately map the lawn, detect obstacles commonly found in yards and green spaces, and provide real-time data for navigation.

We chose a compact yet powerful 3D LiDAR module to give Blix a detection range and field of view suitable for yards between 3000-4000 ㎡ (32,000-43,000 sq ft), letting it work in various environmental conditions. Rain or shine, night or day, chilly spring drafts or summer heat won't bother Blix a bit.

 

A visualization of Blix mapping a lawn.

A combination of LiDAR and IMU (Inertial Measurement Unit) sensors through the SLAM (Simultaneous Localization and Mapping) method provides several key benefits:

  • Accurate Positioning: The LiDAR sensor generates a precise map of the environment, allowing the Blix to determine its location relative to its homebase (the charging station), obstacles, and virtual boundaries.
  • Smart Navigation: The IMU contributes to maintaining the mower’s orientation and stability, ensuring smooth operation even on uneven terrain.
  • Dynamic Obstacle Avoidance: By continuously updating its map and position, the Blix can navigate around obstacles in real-time, enhancing safety and efficiency.
  • Autonomous Path Planning: The combination of these technologies allows the Blix to autonomously plan its mowing routes, ensuring optimal coverage and minimizing the need for manual intervention.

The Role of a Vision Sensor

While LiDAR excels at measuring distances and creating 3D models, vision sensors, or cameras, provide complementary information about the environment. If LiDAR gives the Blix a colorless idea of solid objects in its vicinity, the vision sensor uses color to help it "understand" what every component and structure is. Most importantly, what is grass and what isn't.

There are several types of vision sensors, including RGB cameras, depth cameras, and thermal cameras. Of the wide array of RGB or color cameras, there are also grey-scale stereo cameras, RGB depth cameras, and monocular cameras. Our goal is to use a minimally viable hardware solution in combination with advanced algorithms to enable Blix to comprehend its surroundings with vision.

For Blix, we opted for a monocular RGB camera. At the core of our technology is the application of deep learning algorithms trained on vast datasets. These algorithms allow Blix to perform semantic segmentation, a process that involves classifying each pixel in an image to understand the context of the scene. This is crucial for Blix to recognize and differentiate between various objects in their environment, such as grass, trampolines, playsets, pathways, and flowerbeds.

The more Blix runs on one particular yard, it becomes more adept at understanding its surroundings, and is able to make informed decisions about navigation and task execution. This increased perception not only enhances efficiency but also contributes to overall safety, particularly in dynamic environments where humans or pets might interrupt a task

 

A visualization of what Blix interprets as it mows a lawn.

The Bottom Line - Smart Wireless Robotic Lawn Mower with No RTK

With the combined power of LiDAR and vision, Blix knows what it's doing. Adding on a precise cutting system with 6 razor blades, a collision bumper sensor, a lift sensor, and strong wheel motors, Blix has brains and brawn.

By leveraging these technologies, we ensure precise navigation, effective obstacle avoidance, and a beautifully manicured lawn with minimal effort on your part. With Blix, the idea is to sit back, knowing it's handled each and every time.

Ready to find out for yourself? Visit the Blix product page or leave your information in the form below to stay updated. Transform your lawn care routine with the future of mowing technology!

Bu gönderiyi paylaş
Arşivle
Meet the Sveaverken Blix Robotic Lawn Mower: Loving Your Lawn Has Never Been So Easy