We demonstrate the possibility of learning drone swarm controllers that are zero-shot transferable to real quadrotors via large-scale multi-agent end-to-end reinforcement learning. We train policies parameterized by neural networks that are capable of controlling individual drones in a swarm in a fully decentralized manner. Our policies, trained in simulated environments with realistic quadrotor physics, demonstrate advanced flocking behaviors, perform aggressive maneuvers in tight formations while avoiding collisions with each other, break and re-establish formations to avoid collisions with moving obstacles, and efficiently coordinate in pursuit-evasion tasks. We analyze, in simulation, how different model architectures and parameters of the training regime influence the final performance of neural swarms. We demonstrate the successful deployment of the model learned in simulation to highly resource-constrained physical quadrotors performing station keeping and goal swapping behaviors. Code and video demonstrations are available at the project website https://sites.google.com/view/swarm-rl.

No abstract

Many dental alloy manufacturers instruct not to recast alloys, but the dental laboratories reuse the casting surplus for economic reasons. It is a controversial topic in dental practice, so the present study attempts to assess the effects of recasting Ni-Cr-Mo alloys. Three sets of alloy pellets were recast. The first set was melted and allowed to solidify. The second and third sets were recast two times and three times, respectively. The elemental composition of all the recast samples was analysed using ED-XRF (Energy Dispersive X-ray Fluorescence) spectrometer. The variation in the chemical composition for the number of recasting was reported. It was observed that the recasting for one time doesn't change the elemental composition to a considerable extent. However, further increase in the number of recasting two and three times, the depletion of the major alloying elements (Ni, Cr, and Mo) was notable. Hence, the current study intends to calculate the exact amount of the depleted elements after each recast. Also, to have a theoretical analysis of the elements necessary for making a master alloy that can be added during the recasting to avoid changes in elemental composition. Moreover, the microstructure of the recast samples was observed using an optical electron microscope (OEM). No drastic variation in the microstructure was observed for the alloy with several melting and solidifying cycles, except for the orientation of the dendritic arm. Furthermore, to confirm the mechanical strength of the recast alloy Vickers microhardness test was conducted. The average microhardness of the base Ni-Cr-Mo alloy was 216 HV, and recasting once does not affect the hardness value. However, the three-time recast alloy showed only a 9% decrement. Finally, it can be concluded that the number of recasting can be as many times provided depleted elements are added in exact proportion after each recast. The current research suggests recycling dental alloys in bulk outside a dental clinic, and a dentist should responsibly segregate different base metal alloys and promote sustainable dentistry.

No abstract