Grzegorz Granosik

Autonomous landing capability is crucial for fully autonomous UAV flight. Currently, most solutions use either color imaging from a camera pointed down, lidar sensors, dedicated landing spots, beacons, or a combination of these approaches. Classical strategies can be limited by either no color data when lidar is used, limited obstacle perception when only color imaging is used, a low field of view from a single RGB-D sensor, or the requirement for the landing spot to be prepared in advance. In this paper, a new approach is proposed where an RGB-D camera mounted on a gimbal is used. The gimbal is actively actuated to counteract the limited field of view while color images and depth information are provided by the RGB-D camera. Furthermore, a combined UAV-and-gimbal-motion strategy is proposed to counteract the low maximum range of depth perception to provide static obstacle detection and avoidance, while preserving safe operating conditions for low-altitude flight, near potential obstacles. The system is developed using a PX4 flight stack, CubeOrange flight controller, and Jetson nano onboard computer. The system was flight-tested in simulation conditions and statically tested on a real vehicle. Results show the correctness of the system architecture and possibility of deployment in real conditions.

The growing significance of mobile robots in a full spectrum of areas of life creates new challenges and opportunities in robotics. One critical aspect to consider is energy utilization, as accurate prediction plays a vital role in a robot’s reliability and safety. Furthermore, precise prediction offers economic advantages, particularly for robotic fleets, where energy management systems can optimize maintenance costs and operational efficiency. The following review describes the state of the art of energy usage prediction for different types of mobile robots, highlights current trends, and analyses algorithms’ complexity (in implementation and execution), accuracy, and universality.

It is considered to use a manipulator mounted on a satellite in order to perform active debris removal missions. Space manipulator control system needs to take the dynamic model of the satellite-manipulator system into account because of the influence of the manipulator motion on the position and attitude of the satellite. Therefore, precise modelling of the space manipulator dynamics as well as parameter identification are needed in order to improve the credibility of the simulation tools. In this paper we presented the identification of the flexible-joint space manipulator model based on dynamic equations of motion. Experiments were performed in emulated microgravity environment using planar air bearings. The arbitrarily selected joint-space trajectory was performed by the manipulator’s control system. The experiments were repeated multiple times in order to analyze the identification method sensitivity. The identification is based on Simulink SimMechanics model. Thus, the procedure can be used for any space manipulator without the need of obtaining analytical relations for dynamic equations each time. Including joint flexibility and spring viscous damping in the dynamic model allowed to reflect the experimental measurements better than the reference model could. Identified parameters of the flexible joint have values of the same magnitude as corresponding real system parameters.

The on-ground validation of control systems designed for manipulators working in orbit is very difficult due to the necessity of simulating the microgravity environment on Earth. In this paper, we present the possibilities of utilising the KUKA KUBE test-bed with industrial robots to experimentally verify space systems using hardware-in-the-loop tests. The fixed-base KUKA industrial robot is operated in gravitational environment, while the space system model plant is solved in real time parallel to on-ground experiment. The test-bed measurements are the input of the model plant, and the output of the model is treated as an input for the industrial robot actuation. In the performed experiment, the control system based on the Dynamic Jacobian is validated. The desired point that is reached by the manipulator's end-effector is constant in the simulated environment and moving with respect to the test-bed frame. The position of the space manipulator's end-effector is calculated by evaluating dynamics of the satellite in real-time model. The results show that the control system applied to the KUKA robot works correctly. The measurements from the torque sensors mounted in KUKA robot's joints are in accordance with the simulation results. This fact enhances the possibilities of gravity compensation, thus simulating microgravity environment on the test-bed.

The growing interest in the utilization of Unmanned Aerial Vehicles (UAVs) demands minimizing the costs of robot maintenance, where one of the main aspects relates to energy consumption. This manuscript presents a novel approach to create an energy consumption model for UAVs. The authors prove, based on experimentally collected data using a drone carrying various payloads, that Machine Learning (ML) algorithms allow to sufficiently accurately estimate a power signal. As opposed to the classical approach with mathematical modeling, the presented method does not require any knowledge about the drone’s construction, thus making it a universal tool. Calculated metrics show the Decision Tree is the most suitable algorithm among eight different ML methods due to its high energy prediction accuracy of at least 97.5% and a short learning time which was equal to 2 ms for the largest dataset.

The paper presents the concept of using color template charts for the needs of telemedicine, particularly telediagnosis of the stoma. Although the concept is...

A detailed literature analysis depicts that artificial neural networks are rarely used for the power consumption estimation in the mobile robotics field. Instead, researchers prefer to develop analytical models of investigated robots. This manuscript presents a comparison of mathematical models and non-complex artificial neural networks in energy prediction tasks for differential and skid-steer drive robots which move over various types of surfaces. The results show that both methods could be used interchangeably but AI methods are more universal, do not depend on the kinematic structure of a robot and are tolerant for designers not having a complex knowledge about the system.