AI Based Object Detection and Driveable Area Segmentation (Novavision™)
With our object detection neural network powered by computer vision and AI, Novabot can identify different objects like trees, pets, sprinklers, pools and human beings. The robot avoids stationary obstacles and stops immediately when it sees moving objects, making it is safe for kids and pets. Our state-of-the-art semantic segmentation algorithm enables the robot to differentiate the lawn from other areas so that Novabot™ moves strictly within your lawn, avoiding curbs and driveways.
High-accuracy Localization & Coverage Planning Algorithm
There is no need for our users to set up the perimeter wire. With a high-accuracy localization and navigation algorithm based on multiple sensors, Novabot™ creates a map of the surroundings and updates its planning strategies dynamically to move along a safe and collision-free path. As you change your lawn (i.e. move furniture), Novabot™ relearns its surroundings automatically. Our complete coverage path planning algorithm also enables Novabot™ to cut the lawn effectively and accurately.
Home Surveillance System
Visual object detection can alert you when particular moving objects like human beings or pets enter the lawn. It extends the mower to work as a basic home surveillance camera while it finishes mowing the lawn and is sitting in the charging station. Novabot™ will immediately notify and send images to the user through the App when it detects such objects, giving you real-time intrusion detection. From there, users can return home or tell the APP to ignore it.
High-performance AI board and Easy-to-use App
The 5 TOPS (Tera operations per second) cutting-edge AI chips, combined with our CoreMind algorithms, can quickly process the data and enable the Novabot to react and adapt to complex situations. Our App is easy to use, allowing users to control and get the various status of the robot without learning cost.
Now Get the Super Early Bird Price on Indiegogo！
Obstacle detection by Computer vision (Novavision™)
With obstacle detection by computer vision, Novabot can detect and identify different objects like trees, pets, sprinklers, lawn furniture, pools, writing and people. By evaluating colors, depth, and shape of an object, Novabot learns what you have in your backyard or on your lawn and learns to move around the. Filters are applied to what the robot sees, letting Novabot improve its efﬁciency and create 3D maps of its surroundings. Visual recognition overtime remembers where objects are before seeing them, since it can learn environments and adjust how it moves. Scanning its surroundings, Novabot knows where it is and when to avoid things in its way not requiring GPS.
Why use Visual Recognition?
Applying algorithms to lawn care lets the mower be more accurate in its cutting and detection of sitting and moving objects. It improves system performance and safety by enhancing the autonomous system's ability to avoid accidents. Visual recognition lets the robot react quickly to its changing surroundings and avoid restricted areas. If a ball suddenly passes in front of the mower, visual recognition tells it to immediately stop and readjust. Autonomous lawn mowers need to know the difference between the lawn and concrete to make clear cut edges and create different lawn mower patterns and heights for front and back lawns. Visual recognition also lets the mower see what type of grass you have, giving you ﬂexibility to choose how to cut it. Overall, we use visual recognition to improve accuracy and stability of the Novabot.
Home Security System
Visual recognition can learn to detect your family members and alert you when intruders or unknown objects enter the lawn. Algorithms can remember people’s faces like your phone can and detect sudden movements. This extends the mower to work as a home security system that monitors your lawn while it's mowing and while it's sitting in the charging station. Novabot will immediately notify the user through the smartphone APP that there is suspicious activity giving you real-time intrusion detection.
5TFLOATS Computing Chip
Coremind algorithm and 5TFLOATS Computing Chips are used in many technologies like autonomous-driving cars and computers. 5 Tﬂoats chips are able to process the data fast enough and enable the Novabot to react to any unpredictable situation.