Locomotion is realized through making and breaking contact. State-of-the-art constrained nonlinear model predictive controllers (NMPC) generate whole-body trajectories for a given contact sequence. However, these approaches are computationally expensive at run-time. Hence it is desirable to offload some of this computation to an offline phase. In this paper, we hypothesize that conditioning a learned policy on the locations and timings of contact is a suitable representation for learning a single policy that can generate multiple gaits (contact sequences). In this way, we can build a single generalist policy to realize different gaited and non-gaited locomotion skills and the transitions among them. Our extensive simulation results demonstrate the validity of our hypothesis for learning multiple gaits for a biped robot.

In robotics, designing robust algorithms in the face of estimation uncertainty is a challenging task. Indeed, controllers often do not consider the estimation uncertainty and only rely on the most likely estimated state. Consequently, sudden changes in the environment or the robot's dynamics can lead to catastrophic behaviors. In this work, we present a risk-sensitive Extended Kalman Filter that allows doing output-feedback Model Predictive Control (MPC) safely. This filter adapts its estimation to the control objective. By taking a pessimistic estimate concerning the value function resulting from the MPC controller, the filter provides increased robustness to the controller in phases of uncertainty as compared to a standard Extended Kalman Filter (EKF). Moreover, the filter has the same complexity as an EKF, so that it can be used for real-time model-predictive control. The paper evaluates the risk-sensitive behavior of the proposed filter when used in a nonlinear model-predictive control loop on a planar drone and industrial manipulator in simulation, as well as on an external force estimation task on a real quadruped robot. These experiments demonstrate the abilities of the approach to improve performance in the face of uncertainties significantly.

The most cutting edge robots today can open doors, dance to music, and perform acts of remarkable dexterity. They are also very, very expensive. Tha's why NYU Tandon School of Engineering has designed a budget-friendly robot dog that has many of the same capabilities and agility of its pricier counterparts for only a fraction of the cost. The design is completely free and open-source on GitHub and can be easily 3D printed by labs across the world and assembled using off-the-shelf components. It's even so rugged and durable that the researchers say this robotic pup can be tossed into a backpack and easily carried to conferences or field testing sites.

A research team at the NYU Tandon School of Engineering, with the support of the National Science Foundation's National Robotics Initiative 2.0, is building the foundations of a wireless system that takes advantage of superfast fifth-generation (5G) wireless communications to outsource a mobile robots' artificial intelligence (AI) functions to the edge cloud--the server in the cloud closest to the robot. The collaborators, all of whom are members of the faculty of NYU Tandon's renowned NYU WIRELESS center for telecommunications research, will design manipulation and locomotion algorithms that address some important technical hurdles to making 5G networks a viable bridge between robot and server. Shifting AI capabilities from the robot to a remote server offers tantalizing operational benefits, such as allowing robots to perceive the environment, perform complex operations, and make decisions autonomously, all without incurring major energy and weight costs from onboard computational and power-generation equipment. Comprising Ludovic Righetti, professor in the Departments of Electrical and Computer Engineering and Mechanical and Aerospace Engineering; and Siddharth Garg, Sundeep Rangan and Elza Erkip, professors in the Department of Electrical and Computer Engineering, the team will focus on solving issues of reliability, safety of robotic operation under communication degradation, and scalability to multi-robot systems. The collaboration brings expertise in robotics (Righetti), computer architecture and computation (Garg), wireless networks (Rangan and Erkip), and information theory (Erkip).