EP.5 ROBOT BREAKS 100-METER RUN WORLD RECORD
Robo-version of Usain Bolt, a gnome that helps in the garden, inspection robots and much more.
The ultimate sprinter - meet Cassie the robot 🦿
Cassie, the first bipedal robot to use machine learning to control a running gait on outdoor terrain, completed the 100-meter run in 24.73 seconds. The robot developed by Oregon State robotics professor Jonathan Hurst with funding from DARPA, has ostrich-like knees and operates without cameras or external sensors, essentially functioning as if blind.
Cassie was trained for the equivalent of a full year in a simulation environment, compressed to a week through a computing technique known as parallelization – multiple processes and calculations happening at the same time, allowing Cassie to go through a range of training experiences simultaneously.
The robot has established a Guinness World Record for the fastest 100 meters by a bipedal robot.
Inspection patrol using autonomous robots 🔎
A robot that helps with daily inspection tasks. Energy Robotics has installed the robot in the Austrian Alps, in the hydroelectric power plant. The robot is equipped with advanced sensors and camera systems that make the robot 'see' a lot. The main task is to monitor the status of different kinds of gauges with the help of cameras and AI algorithms.
The robot uses thermal images to detect defects in pipes and temperature levels. The collected data can be used not only to detect anomalies such as excessive temperature, leaks, or contamination at an early stage but also to read outflow rates and pressure values. It maneuvers through narrow corridors and climbs and descends stairs with ease while recording and transmitting information about the operation on site.
A robot helper in your garden (it’s not a gnome) 🍃
Is it a robot or is it a gnome? We don't know that 100%, but we do know one thing. This solution looks like it was ripped out of the future. Your personal assistant in the garden. This robot can help you with planting plants, mowing grass, and picking up trash, and will also gladly take on the task of sweeping the sidewalk.
Are we experiencing a moment when the price of robots is no longer astronomical and we will see them more often in relation to what they offer at the other end? Let's see, but Willow is certainly a great forecast!
A throw for three points, but it goes in every time 🏀
The AI basketball robot CUE was created by the Toyota Engineering Society in 2017. It can shoot a maximum of 20m by using high-output actuators, batteries, and technology for analyzing dynamic movements. CUE debuted in 2018 and has evolved with autonomous movement and arms control. It uses a Gaussian process regression model to shoot and Bayesian optimization to find the optimal shooting form. CUE also uses a distance image camera to predict the trajectory of the ball when dribbling.
In 2019, CUE3 achieved a Guinness World Record for the most consecutive free throws made by a humanoid robot. Furthermore, the robot has the ability to improve its skills over time by learning from its mistakes and adjusting the shooting form accordingly. The robot is also capable of playing with a team of human players, which is a significant achievement in the field of robotics.
Robotics term of the week
Vision systems in robotics
Vision systems are used in robotics for several reasons. One of the main reasons is perception, as vision systems allow robots to perceive and understand their environment, which is essential for tasks such as navigation, manipulation, and object recognition. Another reason is accuracy, as vision systems provide a high level of accuracy when measuring distances, sizes, and positions of objects, which is important for tasks such as grasping and manipulation. There are several types of vision systems that are commonly used in robotics, including:
Machine Vision: This type of vision system uses cameras, sensors, and software to interpret and analyze images in order to extract information and make decisions.
Computer Vision: This type of system uses algorithms and techniques from computer science to interpret and understand visual information from the environment.
Stereo vision: This type of system uses two or more cameras to capture images of the same scene and then uses the differences between the images to calculate the distance to objects in the environment.
LiDAR: LiDAR (Light Detection and Ranging) is a sensor technology that uses laser light to measure distances between objects.
RGB-D Camera: RGB-D cameras can provide both color and depth information of the same scene, which can be used in robotics for various tasks such as object recognition, 3D scanning, and navigation.
Thermal Imaging: This type of system uses infrared cameras to capture images of the environment based on the temperature of objects, which can be useful for tasks such as navigation and object detection in low-light or dark conditions.
Multi-modal Vision Systems: These systems combine different types of vision systems to take advantage of the strengths of each. For example, a robot might use LiDAR for precise navigation and stereo vision for grasping objects.