In the rapidly developing era of artificial intelligence (AI), the integration of AI with agriculture occupies a central place. Among the latest ecoorobotix innovations, seven feet, a supported GPS unit powered by solar energy, elegantly shifts crop fields, directing and eliminating weeds with an amazing 95% accuracy, effectively reducing waste. In addition, energetic, universal robots radically change the collection of citrus fruit by combining many cameras and flexible robotic arms. River Setucebot uses scanning of crop geometry to optimize growth and minimize pesticide consumption, distinguishing weeds and crops to prevent shipping and diseases.
However, the current challenge is to move around complex, constantly changing natural environments, such as dense forests or high fields of grass. How can robots effectively remember where there were and recognize places they visited earlier in a visually repetitive environment?
The inspiration was found in an unlikely source: ants. These little creatures show extraordinary navigation skills despite their relatively simple sensory and neural systems. Scientists, led by Le Zhu at the universities in Edinburgh and Sheffield, tried Imitate ants navigation efficiency in a new artificial neural network. This network would help robots recognize and remember routes in complex natural environments, especially in agriculture, where dense vegetation is a significant challenge.
Annts use a unique neural structure known as “mushroom bodies” in their brains to detect visual patterns and storage memories of space -time, enabling them to have a visually effective repetitive environment. ZHU and his team used this biological mechanism as an inspiration for their research.
Their approach consisted of designing a bio -inspired camera camera mounted on a ground robot to capture visual sequences along routes in natural external environments. To facilitate the recognition of routes, they have developed a neuronal algorithm for space -time memory, which strictly reflects the circumference of the body of insect fungi.
Most importantly, they used neuromorphic calculations, imitating the structure and function of biological neurons, to coding memory in a thrilling neuron network operating on a low -power neuromorphic computer. The result was a robotic system that could assess visual knowledge in real time based on material from the camera camera, supporting the recognition of the route for visual navigation.
During strict tests in various settings, including green lands, forests and arable fields, an ant -inspired anthon model proved its effectiveness. This assessed the next method of learning the route called Seqslam when assessed on repeated travel on the same route or routes with small side shifts. Seqslam is a technique that corresponds to the image sequences to find similarities between various mileage.
The implications of these studies go far beyond robotics. This neural model inspired by an ant is a promise to transform agricultural robotics, making it more efficient and effective in navigation by dense vegetation. In addition, scientists suggest that the principles of this model can be extended to other sensory methods, such as olfaction or sound, increasing the perception of its environment by the robot.
This study is a significant step forward in the use of the collective wisdom of nature navigators to increase our technological progress. When we still draw inspiration from the world of nature, robotics based on artificial intelligence can find even more innovative solutions to complex challenges, which ultimately bring wide industries.
Read more most important news about robot navigation here.
