Observing social/physical distancing norms between humans has become an indispensable precaution to slow down the transmission of COVID-19. We present a novel method automatically detect pairs in crowded scenario who are not maintaining social distancing, i.e. about 2 meters space them using autonomous mobile robot and existing CCTV (Closed-Circuit TeleVision) cameras. The is equipped with commodity sensors, namely RGB-D (Red Green Blue—Depth) camera 2-D lidar breaches within their sensing...

We present a novel method for reliable robot navigation in uneven outdoor terrains. Our approach employs fully-trained Deep Reinforcement Learning (DRL) network that uses elevation maps of the environment, pose, and goal as inputs to compute an attention mask environment. The is used identify reduced stability regions map computed using channel spatial modules reward function. continuously update cost-map encodes information or level-of-flatness terrain mask. then generate locally least-cost...

We present Frozone, a novel algorithm to deal with the Freezing Robot Problem (FRP) that arises when robot navigates through dense scenarios and crowds. Our method senses explicitly predicts trajectories of pedestrians constructs Potential Zone (PFZ); spatial zone where could freeze or be obtrusive humans. formulation computes deviation velocity avoid PFZ, which also accounts for social constraints. Furthermore, Frozone is designed robots equipped sensors limited sensing range field view....

We present DenseCAvoid, a novel algorithm for navigating robot through dense crowds and avoiding collisions by anticipating pedestrian behaviors. Our formulation uses visual sensors trajectory prediction to track pedestrians in set of input frames compute bounding boxes that extrapolate the positions future time. hybrid approach combines this with Deep Reinforcement Learning-based collision avoidance method train policy generate smoother, safer, more robust trajectories during run-time. our...

Maintaining social distancing norms between humans has become an indispensable precaution to slow down the transmission of COVID-19. We present a novel method automatically detect pairs in crowded scenario who are not adhering distance constraint, i.e. about 6 feet space them. Our approach makes no assumption crowd density or pedestrian walking directions. use mobile robot with commodity sensors, namely RGB-D camera and 2-D lidar perform collision-free navigation estimate all detected...

We present a novel Deep Reinforcement Learning (DRL) based policy to compute dynamically feasible and spatially aware velocities for robot navigating among mobile obstacles. Our approach combines the benefits of Dynamic Window Approach (DWA) in terms satisfying robot's dynamics constraints with state-of-the-art DRL-based navigation methods that can handle moving obstacles pedestrians well. formulation achieves these goals by embedding environmental obstacles' motions low-dimensional...

We present a novel high fidelity 3-D simulator that significantly reduces the sim-to-real gap for collision avoidance in dense crowds using Deep Reinforcement Learning (DRL). Our models realistic crowd and pedestrian behaviors, along with friction, sensor noise delays simulated robot model. also describe technique to incrementally control randomness complexity of training scenarios achieve better convergence generalization capabilities. demonstrate effectiveness our by policy fuses data from...

We present a novel trajectory traversability estimation and planning algorithm for robot navigation in complex outdoor environments. incorporate multimodal sensory inputs from an RGB camera, 3D LiDAR, the robot's odometry sensor to train prediction model estimate candidate trajectories' success probabilities based on partially reliable multi-modal observations. encode high-dimensional low-dimensional feature vectors using encoder networks represent them as connected graph. The graph is then...

We present a novel learning-based collision avoidance algorithm, CrowdSteer, for mobile robots operating in dense and crowded environments. Our approach is end-to-end uses multiple perception sensors such as 2-D lidar along with depth camera to sense surrounding dynamic agents compute collision-free velocities. training based on the sim-to-real paradigm high fidelity 3-D simulations of pedestrians environment train policy using Proximal Policy Optimization (PPO). show that our learned...

We present CoMet, a novel approach for computing group's cohesion and using that to improve robot's navigation in crowded scenes. Our uses cohesion-metric builds on prior work social psychology. compute this metric by utilizing various visual features of pedestrians from an RGB-D camera on-board robot. Specifically, we detect characteristics corresponding the proximity between people, their relative walking speeds, group size, interactions members. use our design scheme accounts different...

Aluminum 7075 alloy (AA 7075) is one of the prime materials used in aviation and automotive industry because its high strength to weight ratio, good amount fatigue machinability. Friction stir processing (FSP) emerging solid state process that refines microstructure hence improved mechanical properties are obtained. The temperature during FSP affects resulting so attempt for reducing can result into reduction grain size. fine size delivers percentage elongation which reduces number joints...

We present a novel Deep Reinforcement Learning (DRL) based policy to compute dynamically feasible and spatially aware velocities for robot navigating among mobile obstacles. Our approach combines the benefits of Dynamic Window Approach (DWA) in terms satisfying robot's dynamics constraints with state-of-the-art DRL-based navigation methods that can handle moving obstacles pedestrians well. formulation achieves these goals by embedding environmental obstacles' motions low-dimensional...

Due to the rapid expansion of infrastructure, demand for Uninterruptible Power Supply (UPS) is gradually increasing. For good redundancy and less downtime, parallel operation (N+1) a special feature high performance UPS system. The main objective this configuration operate all UPSs independently. challenging task as synchronization among connected inverters has be maintained. In work, Master-Slave (M-S) based concept used N+1 problem associated with operation, such unequal load sharing...

We present a novel trajectory traversability estimation and planning algorithm for robot navigation in complex outdoor environments. incorporate multimodal sensory inputs from an RGB camera, 3D LiDAR, the robot's odometry sensor to train prediction model estimate candidate trajectories' success probabilities based on partially reliable multi-modal observations. encode high-dimensional low-dimensional feature vectors using encoder networks represent them as connected graph. The graph is then...

This paper describes the basic concepts of design, assembly, and testing HVAC electrical panels.Here are complete details about components that used in designing panel.HVAC panel consists GA drawing panel, a power circuit for system, control system.After considering different parameters, assembly will cover up as per guidelines requirements consumer.

Users are favoring of distributed uninterruptible power supply (UPS) systems as an alternative to a large conventional UPS system. To meet expectations the users, parallel operation is special feature systems. In present work master-slave based configuration used for connected and problem associated with sharing addressed. The aim overcome challenges unequal by using synchronous reference frame harmonic control loops. proposed topology load current inverter information each unit achieve...