Video Friday is a weekly selection of amazing robotics videos collected by your friends on IEEE Spectrumrobotics. We also publish a weekly calendar of upcoming robotics events for the next few months. Please send us your events for inclusion.

HRI 2023: March 13-16, 2023, STOCKHOLM
Robotics Summit & Expo: May 10-11, 2023, BOSTON
ICRA 2023: May 29 – June 2, 2023, LONDON
RoboCup 2023: July 4-10, 2023, BORDEAUX, FRANCE
RSS 2023: July 10-14, 2023 DAEGU, KOREA
IEEE RO-MAN 2023: August 28-31, 2023, BUSAN, KOREA
CLAWAR 2023: October 2-4, 2023, FLORIANOPOLIS, BRAZIL
Humanoids 2023: December 12-14, 2023, AUSTIN, TX.

Enjoy today’s videos!

Agilicious is a jointly developed hardware/software framework for autonomous and maneuverable quadcopter flight that has been developed and used since 2016 at the Robotics and Perceptions Group of the University of Zurich. Agilicious is completely open source hardware that supports both model based and neural network based controllers!

[ Agilicious ]

The adaptive Flexiv massage robot combines force control, computer vision and artificial intelligence to emulate the skills and agility of a human massage therapist.

[ Flexiv ]

Loosely inspired by “The Wolf and the Lamb” by Jean de la Fontaine, the Coperni fashion show in Paris. shows Spot robots from Boston Dynamics in relation to humans. Unlike the original fable, written in the 17th century, which raises questions regarding the balance of power between the human groups that make up society, Coperni reimagines history and brings it to the year 2023 with a positive vision of the future. The wolf figure is replaced by Spot robots, and the lambs are replaced by humans. The fashion show tells the story of their meeting, their mutual domestication and living together. The show presents Coperni’s vision, which is that there is neither dominant nor dominated, but that humanity and the machine live in harmony.

[ Coperni ]

Meet the third-generation Honda Autonomous Work Vehicle (AWV) prototype, a new category of work vehicle that can operate autonomously in a variety of dynamic work environments. The Honda AWV can improve efficiency, increase productivity and improve environmental performance in the construction industry and other industries that require an autonomous off-road solution.

[ Honda ]

Skydio’s partnership with WeRobotics provides micro-grant scholarships to Flying Labs around the world, with Nepal Flying Labs selected as the recipient for its project to create a digital twin of Changu Narayan Temple, the oldest temple in the Kathmandu Valley dating back to the 5th century.

[ Skydio ]

This is perhaps the most brisk walk I have ever seen on a humanoid robot.

[ GitHub ]

An interesting “autoloader” for Wing delivery drones.

[ Wing ]

In this video, we show how robots can learn from human experts to master the complex sequence of tasks in various manufacturing processes such as surface treatment and composite laying. We demonstrate how robots can learn expert task sequence policies for a complex stacking task and perform execution on a completely new test piece.

[ USC Viterbi ]

We present a learning-based approach based on the simulation of real humanoid movements. We do not use state estimation, dynamics models, trajectory optimization, reference trajectories, or precomputed gait libraries. Our controller is trained with large-scale modelless reinforcement learning in an ensemble of randomized environments in simulation and deployed in the real world in zero-shot mode. To the best of our knowledge, this is the first demonstration of a fully learning-based method of locomotion for full-sized humanoids in the real world.

[ Paper ]

Roboball? Roboball.

[ Texas A&M ]

In this paper, the problem of active co-localization of robot commands with unknown data association is considered. Specifically, it involves positioning a small number of identical unmanned ground vehicles (UGVs) at desired positions so that the unmanned aerial vehicle (UAV) can unambiguously determine its global position through untagged UGV measurements. This work provides robust localization in difficult-to-read environments where GPS is not available, thus paving the way for large-scale multi-robot navigation and mapping.

[ UPenn ]

This paper presents a multi-segmented vine robot that can navigate complex paths without interacting with the environment. This is achieved through a new control method that selectively actuates each individual bag at the tip, providing a high degree of freedom with a small number of control inputs.

[ Paper ]

TBNET and DEEP Robotics have deployed the Jueying X20 four-legged robot with two high-precision surveying and mapping devices. He completed field data collection in the complex terrain of real scenes such as indoor and outdoor construction sites, electrical pylon clusters, railways, underground garages, and ancient buildings.

[ DeepRobotics ]

Here is a brief overview of RoMeLa’s new robotic limb EEMMMa (Elastic Stretch Mechanism for Mobility and Manipulation), a long steel tentacle that can both stretch and bend. It can be used to deploy climbing hooks and change its shape to avoid obstacles. This allows cameras to be placed and samples to be taken from hard-to-reach places. The unique limb spring design provides a versatile, lightweight and compact system. This could allow future mobile robots to easily and safely navigate highly unstructured terrain such as forests or cave systems.

[ RoMeLa ]

Microrobots are a revolutionary new technology that could change the way we interact with the world around us. For the first time, a joint research team of electrical and computer scientists, backed by NSF, has installed electronic brains on solar-powered microbots the size of a human hair. One of the biggest problems is their small size, requiring external control such as a computer or smartphone, which limits their range and makes remote control of bots difficult until now.

[ NSF ]

A team of scientists, engineers and designers is sent on an Arctic expedition to test space technology. The MIT Space Research Initiative expedition to Svalbard was not just a space analog mission, but an experience to learn how to improve access to remote regions from the far reaches of planet Earth, to the Moon and Mars.

[ MIT ]

The Perseverance rover, which landed on Mars in February 2021, has the most advanced autonomous driving capabilities ever flown to Mars. Having such advanced capabilities contributes to the rover’s challenging mission to detect signs of life that may have existed on Mars in the distant past. This paper provides an overview of current research and development in robotics autonomy at JPL, with a focus on improving the safety, efficiency, and productivity of robot mobility through risk-based decision-making and machine learning applications.

[ JPL ]

An interactive webinar discussing the possibilities and path to general robot manipulation. This will be the first event in a series organized by the Northwestern Center for Robotics and Biosystems and dedicated to this daunting task of robotics.

[ Northwestern ]