Recent advancements in large-scale offline training have demonstrated the potential of generalist policy learning for complex robotic tasks. However, applying these principles to legged locomotion remains a challenge due to continuous dynamics and the need for real-time adaptation across diverse terrains and robot morphologies. In this work, we propose GRoQ-LoCO, a scalable, attention-based framework that learns a single generalist locomotion policy across multiple quadruped robots and terrains, relying solely on offline datasets. Our approach leverages expert demonstrations from two distinct locomotion behaviors - stair traversal (non-periodic gaits) and flat terrain traversal (periodic gaits) - collected across multiple quadruped robots, to train a generalist model that enables behavior fusion. Crucially, our framework operates solely on proprioceptive data from all robots without incorporating any robot-specific encodings. The policy is directly deployable on an Intel i7 nuc, producing low-latency control outputs without any test-time optimization. Our extensive experiments demonstrate zero-shot transfer across highly diverse quadruped robots and terrains, including hardware deployment on the Unitree Go1, a commercially available 12kg robot. Notably, we evaluate challenging cross-robot training setups where different locomotion skills are unevenly distributed across robots, yet observe successful transfer of both flat walking and stair traversal behaviors to all robots at test time. We also show preliminary walking on Stoch 5, a 70kg quadruped, on flat and outdoor terrains without requiring any fine tuning. These results demonstrate the potential of offline, data-driven learning to generalize locomotion across diverse quadruped morphologies and behaviors.

In recent years, quadruped robots have attracted significant attention due to their practical advantages in maneuverability, particularly when navigating rough terrain and climbing stairs. As these robots become more integrated into various industries, including construction and healthcare, researchers have increasingly focused on developing intuitive interaction methods such as speech and gestures that do not require separate devices such as keyboards or joysticks. This paper aims at investigating a comfortable and efficient interaction method with quadruped robots that possess a familiar form factor. To this end, we conducted two preliminary studies to observe how individuals naturally interact with a quadruped robot in natural and controlled settings, followed by a prototype experiment to examine human preferences for body-based and hand-based gesture controls using a Unitree Go1 Pro quadruped robot. We assessed the user experience of 13 participants using the User Experience Questionnaire and measured the time taken to complete specific tasks. The findings of our preliminary results indicate that humans have a natural preference for communicating with robots through hand and body gestures rather than speech. In addition, participants reported higher satisfaction and completed tasks more quickly when using body gestures to interact with the robot. This contradicts the fact that most gesture-based control technologies for quadruped robots are hand-based. The video is available at https://youtu.be/rysv1p1zvp4.

A recent study found robots that speak in a "charismatic" tone while directing a college class can boost creativity among humans. Artificial intelligence could soon start replacing household pets -- no vet bills required. As various types of robots continue to hit the market, AI-powered "animals" have arrived on the scene as well. One example is Go1, the world's first intelligent quadruped robot "companion" that is developed by China's Unitree Robotics. The robotic sidekick walks on all fours, much like a dog -- but there's no need for a collar or a leash.

U.S.-based Boston Dynamics' Spot is distinguishedly the market leader regarding these robots. Interesting Engineering has previously reported on many of Spot's antics and cuteness. Through videos such as the one shared above, Unitree is also looking to pique the cuteness quotient of its offerings. However, there are many other reasons why one could pick a Unitree robotic dog. Unitree's robotic dog, Go1, does not boast bright colors and only has a silvery metallic appearance.

You're probably familiar with Spot, Boston Dynamics' highly advanced, nightmare-inducing robot dog. And while it went on sale last year, few of us have an extra $74,500 lying around to buy one. However, Chinese firm Unitree Robotics has a similar quadruped bot that's not only a fraction of the size, but it also starts at a mere $2,700. For an advanced robot dog, that's actually pretty dang affordable. In a video, you can see the bot walking alongside its "owner" while also automatically avoiding obstacles in its path.

Boston Dynamics isn't the only company that makes futuristic quadrupedal robots. Chinese firm Unitree Robotics has also been at it for years, and this week revealed its latest creation: the Unitree Go1, a robust-looking four-legged bot that's remarkably cheap, with prices starting at just $2,700. What is the Go1 for, though? Well, a demo video shows it being put to such useful tasks as "following someone on a run" and "carrying a single bottle of water." Sure it's not practical to have a robot butler for your phone and wallet, but it makes a statement on a night out. More realistically, the robotics industry is still exploring the best applications for these sorts of machines.

As unions, corporations and governments debate what effect artificial intelligence, machine learning and automation will have on the future of the workforce, venture capital investors are identifying the most interesting start-up investments that may help steer this historic paradigm shift. Among them is Microsoft's M12 venture fund and Dell Technologies Capital. Both aim to advance human progress and focus on start-ups they can mentor with their companies' own technical expertise and market know-how. "Our goal is to get a window on innovation," says Scott Darling, president of Dell Technologies Capital, who notes that Michael Dell reviews every single deal the fund invests in. "We need to plug into the external entrepreneurial ecosystem. This is so important, since the pace of technology is stunning."