(Web Desk) - Researchers have developed a specialized autonomous robot to remove waste particles and help restore environmental balance.
A team at the Italian Institute of Technology (IIT) in Genoa has created a novel platform, VERO (Vacuum-cleaner-equipped RObot), to help further this noble initiative.
The robot helps people to clean parks, beaches, and narrow alleys—especially when it comes to getting rid of cigarette butts—in areas that are difficult for wheeled and tracked robots to access.
The four legs are connected to the onboard vacuum cleaner using hoses. Each foot has a vacuum cleaner nozzle attached, which enables the robot to effectively gather cigarettes while it is moving by just walking up to the thing it detects and turning on the vacuum.
Litter poses a significant threat to the balance of many ecosystems, particularly in marine environments. Coastal and urban litter travels through gutters, streets, and waterways, eventually reaching the sea.
As it decomposes, this litter releases toxic chemicals and microplastics, further polluting the environment. Traditionally, litter removal relies on manual labor, which limits the amount of waste that can be effectively collected.
The team introduces an innovative quadruped robot prototype designed to autonomously collect cigarette butts (CBs) – the second most common litter item globally. It is estimated that 4.5 trillion cigarettes are disposed of in the environment every year.
Researchers claim that with its natural mobility, the robot can navigate and clean terrains that are inaccessible to wheeled and tracked robots.
The prototype is built on the AlienGo platform from Unitree, equipped with a commercial vacuum attached to its back. Custom 3D-printed nozzles are fitted to the end of hoses that run from the vacuum down each leg to the feet, maximizing ground-level suction without hindering the robot’s movement, reports IEEE Spectrum.
What is truly novel is that the system uses a convolutional neural network to spot litter and then employs a planner to quickly figure out the best way to gather all the detected objects.
For precise cleanup, a visual-servoing system guides the vacuum nozzle, attached to one of the robot’s legs, directly over the identified cigarette butt.
Visual-servoing is a technique that uses feedback information extracted from a vision sensor to control the motion of a robot.
Tests showed that the robot is capable of performing the collection task without stopping its motion, significantly increasing the time efficiency of the entire process.
Extensive testing was conducted in six different outdoor scenarios to demonstrate the prototype’s and method’s performance.
In initial tests across various environments, the robot successfully collected nearly 90 percent of cigarette butts. This efficiency likely surpasses human performance, as humans can lose patience. Despite its slower pace, the robot’s ability to work continuously as long as its battery lasts makes speed less critical.
The team asserts that this is the first instance where a legged robot’s legs are used simultaneously for movement and another task. Unlike other robots that temporarily repurpose their feet as manipulators to perform actions like opening doors, this robot continues to use its legs for locomotion while carrying out additional tasks, reports