Reward signals in reinforcement learning are expensive to design and often require access to the true state which is not available in the real world. Common alternatives are usually demonstrations or goal images which can be labor-intensive to collect. On the other hand, text descriptions provide a general, natural, and low-effort way of communicating the desired task. However, prior works in learning text-conditioned policies still rely on rewards that are defined using either true state or labeled expert demonstrations. We use recent developments in building large-scale visuolanguage models like CLIP to devise a framework that generates the task reward signal just from goal text description and raw pixel observations which is then used to learn the task policy. We evaluate the proposed framework on control and robotic manipulation tasks. Finally, we distill the individual task policies into a single goal text conditioned policy that can generalize in a zero-shot manner to new tasks with unseen objects and unseen goal text descriptions.

This paper details a method for identifying a set of piece-wise affine linear models that can be used for control design for flapping-winged flight. The paper focuses on diving maneuvers as the application for these models. The flight conditions during the dive are segmented into separate dynamically similar regions, and least-squares is used to estimate affine linear models for each modeling region. These models are used to compute the reachability sets that satisfy recovery conditions for safe diving. The point within the dive to begin recovery was determined by checking the current pose for inclusion in the backward reachable set. Using this control method, 2.2 meter dives were achieved at a success rate of 60 percent. The data-driven automatic modeling techniques and controller design processes can be extended to additional flight maneuvers, provided sufficient previous data have been collected for model generation of those maneuvers.

In this work, we develop a cooperative launching system for a 13.2 gram ornithopter micro-aerial vehicle (MAV), the H2Bird, by carrying it on the back of a 32 gram hexapedal millirobot, the VelociRoACH. We determine the necessary initial velocity and pitch angle for take off using force data collected in a wind tunnel and use the VelociRoACH to reach these initial conditions for successful launch. In the wind tunnel predicted success region, we were able to complete a successful launch for 75 percent of the 12 trials in that region. Although carrying the H2Bird on top of the VelociRoACH at a stride frequency of 17 Hz increases our average power consumption by about 24.5 percent over solo running, the H2Bird, in turn, provides stability advantages to the VelociRoACH. We observed that the variance in pitch and roll velocity with the H2Bird is about 90 percent less than without. Additionally, with the H2Bird flapping at 5 Hz during transport, we observed an increase of 12.7 percent of the steady state velocity. Lastly, we found that the costs of transport for carrying the H2Bird flapping and without (6.6 and 6.8) are lower than the solo costs of transport for the VelociRoACH and for the H2Bird (8.1 and 10.1).