Cooperative multi-agent aerial systems are transforming aerial logistics and manipulation by enabling drones to perform dynamic physical interactions. This paper presents the complete modeling and control of a two-quadcopter system executing a high-speed throw-and-catch maneuver. A high-fidelity MATLAB simulation couples a throwing unmanned Aerial Vehicle (UAV) that launches a ballistic payload with a catching UAV that autonomously intercepts it. The main challenge lies in achieving reliable interception of a fast, unpowered projectile under severe dynamic and latency constraints. Conventional reactive-pursuit controllers, which continuously track the projectile, were found to be unstable and computationally infeasible. To overcome these limitations, a lightweight Predictive Intercept Controller (PIC) is proposed, which computes a single fixed spatio-temporal intercept point using the projectile’s ballistic model and commands a tuned Proportional–Derivative (PD) controller to reach it. Monte Carlo evaluations demonstrate a 93.3% interception success rate across randomized launch conditions, with mean spatial and temporal errors below 0.12 m and 40 ms, respectively. The results confirm that the proposed predictive–feedback framework is robust, repeatable, and suitable for real-time implementation on embedded UAV platforms.

Planetary bodies characterized by low gravitational acceleration, such as the Moon and near-Earth asteroids, impose unique locomotion constraints due to diminished contact forces and extended airborne intervals. Among traversal strategies, hopping locomotion offers high energy efficiency but is prone to mid-flight attitude instability caused by asymmetric thrust generation and uneven terrain interactions. This paper presents an underactuated bipedal hopping robot that employs an internal reaction wheel to regulate body posture during the ballistic flight phase. The system is modeled as a gyrostat, enabling analysis of the dynamic coupling between torso rotation and reaction wheel momentum. The locomotion cycle comprises three phases: a leg-driven propulsive jump, mid-air attitude stabilization via an active momentum exchange controller, and a shock-absorbing landing. A reduced-order model is developed to capture the critical coupling between torso rotation and reaction wheel dynamics. The proposed framework is evaluated in MuJoCo-based simulations under lunar gravity conditions (g = 1.625 m/s^2). Results demonstrate that activation of the reaction wheel controller reduces peak mid-air angular deviation by more than 65

Obstacle crossing is an important ability in biped and humanoid robots that are designed to traverse unstructured terrain. We consider the problem of determining the maximum (a) height, (b) width, (c) cross-sectional area, (d) thin vertical barrier height, and (e) square area of the obstacle that an underactuated biped robot with point-feet can cross while walking slowly. Two different biped robot configurations are compared for obstacle crossing: revolute knee and prismatic knee. The path needed to overcome the obstacle without touching it is determined with the help of binary occupancy grid in the sagittal plane and using genetic algorithm based maximization for each of the five cases, considering thin links as well as thick links for the biped robots. The determined collision free path for obstacle crossing is implemented as a trajectory and demonstrated in dynamic simulation in Mujoco simulation environment. In order to control the position of zero moment point (ZMP) and the ground projection of center of mass for stability, a reaction wheel in the torso is utilized. It is observed that increasing the thicknesses of the biped robot links in general has an effect of reducing the maximum size of the obstacle that can be crossed. Further, prismatic knee biped robot performs better than revolute knee biped robot in crossing large obstacles, especially with thick links. Experiments on a prismatic-knee biped robot further validate the results of GA and MuJoCo simulations.

This paper presents a novel aerial-ground robotic system that integrates a hexacopter UAV with a vibration-isolated quadruped robot for autonomous deployment, environmental interaction, and delivery of objects in unstructured terrains. We introduce a unique charpai-inspired spring-thread suspension platform. This passive system effectively attenuates aerial vibrations during flight, enabling stable in-flight transport of a 12-DOF quadruped. The robot employs a front-mounted, ant-inspired 2-DOF gripper and a ROS2-based control stack for vision-guided navigation and inertial stabilization. Unlike prior UAV-UGV platforms, our system enables active mid-air balancing, terrain-aware deployment, and autonomous object manipulation, validated in complex outdoor environments. Theoretical modeling and experimental validation reveal over 85% vibration isolation efficiency, with field trials confirming robust performance in complex outdoor environments. This work establishes a scalable and robust hybrid framework enabling autonomous monitoring, disaster response, and autonomous precision delivery in challenging terrains.

Personalized vehicles have gained significant popularity in urban environments due to their simplicity and convenience. Among these, unicycles are emerging as one of the preferred choice for short-distance travel. However, the unique dynamics of unicycles present challenges related to ride comfort and safety, particularly when they are exposed to vibrations from uneven road surfaces. Understanding the biodynamic responses of riders under such conditions is critical for optimizing the unicycle design and ensuring compliance with vibration exposure standards. This study investigates the biodynamic responses of a unicycle rider subjected to whole-body vibrations across various road surfaces. The human body is modelled as a lumped parameter system with 13 degrees of freedom (DOF), where the unicycle is modelled analogous to a quarter-car system. The equations of motion were derived using the Euler-Lagrange method, and the biodynamic responses were evaluated in accordance with the ISO 5982-2001 standard. Key metrics such as Foot-to-Head Transmissibility (FTHT), Driving Point Mechanical Impedance (DPMI), and Apparent Mass (AM) are derived and computed to evaluate the vibration behaviour of the system. In addition, the road surfaces were modelled following the ISO 8606 standard, and the effects were incorporated into the rider-system dynamics. The peak acceleration values for individual body segments were analyzed for different road conditions, highlighting the pelvis and thighs as the segments experiencing the highest vertical vibration transmission. In addition, the primary response of the system was observed at approximately 9-10 Hz, which can be attributed to the vibration modes of the thigh and pelvic pitches.

Prismatic-knee biped robots have recently been considered an alternative to revolute-knee biped robots. In this study, we investigate control and stabilization strategies for a three-link biped robot with point feet, actuated at the hip and knee joints, making it an underactuated system. Two architectural configurations are considered: (i) a model with a prismatic knee joint, and (ii) a model with a conventional revolute knee joint. The investigation aims to evaluate and compare how these two knee designs affect the robot’s ability to achieve dynamic stability, especially when subjected to diverse disturbances or non-equilibrium initial configurations. For both the biped robot variants, the work space of the hip joint is systematically sampled to generate a set of disparate initial conditions. These scenarios reflect practical situations where the robot may encounter substantial deviations from its standing equilibrium–such as after a push, slip, or terrain irregularity. The control policy is trained using the Soft Actor-Critic (SAC) algorithm in a custom Gymnasium environment built with the MuJoCo simulator using default frictional values. The observation space includes ground reaction forces, joint velocities, joint positions, inertia matrix, center of mass velocities.

This paper presents a reinforcement learning (RL) framework for controlling a planar bipedal robot with nine degrees of freedom (DOF), incorporating prismatic joints in both the shank and thigh segments. The inclusion of prismatic knee and thigh joints allows the robot to dynamically adjust its leg length during locomotion, significantly enhancing adaptability to uneven terrain and improving walking stability. The control architecture is implemented within a custom Gymnasium environment, leveraging the MuJoCo physics engine for high-fidelity simulation of the robot’s dynamics, contact interactions, and frictional effects. The RL policy is trained using state-of-the-art algorithms, with observation inputs comprising joint positions and velocities, ground reaction forces, and friction estimates. Simulation results demonstrate that the bipedal robot can successfully navigate uneven terrain without falling throughout the evaluated episode length, while ensuring all joint torque commands remain within the specified actuator limits.

This paper presents the design and development of an aerial manipulator utilizing a hybrid arm-leg system for enhanced aerial manipulation and mobility. The objective is to investigate the feasibility of utilizing a 4-degree-of-freedom (4DOF) hybrid manipulator consisting of two 2-DOF arms to maintain stable drone hovering while performing grasping tasks. The experiments were conducted for the effectiveness of the proposed method in achieving stable flight with objects, ensuring the avoidance of object loss or dropping. The integration of arms and legs in the hybrid manipulator provides increased versatility and adaptability, enabling agile and precise grasping tasks in challenging environments. Prospective developments could involve refining control algorithms and investigating novel materials and designs to enhance the manipulator's precision, efficacy, and overall functionality.

This article presents the limit-cycle walking of a three-link biped robot consisting of a torso, two knee-locked legs, and point feet. The robot has three degrees of freedom, with its actuated hip joints and unactuated toe joints. The study proposes a PD feedback controller based on trajectory optimization, which is derived mathematically from Lyapunov analysis. The controller parameters are optimized using a genetic algorithm. The paper also tests the robustness of the model by applying external disturbances and shows that the Proportional Derivative controller enhances the robot's robustness, even in the presence of significant external disturbances.

This paper presents the RL framework for walking of a 6DOF planar bipedal robot with a prismatic knee joint. The prismatic knee joint allows the robot to adapt to varying terrain and maintain stability during dynamic movements. The design is based on a two-legged walking model, where each leg has a prismatic knee joint and two other joints: hip and ankle are revolute. The control we use is Reinforcement Learning based system that utilizes an Open AI Gym MuJoCo environment and stable-baselines-3 as a tool to train the Biped model. The friction, ground reaction forces, velocity and position are considered as observation space for neural network input. The robot was tested for walking on the plane floor and its performance was compared to that of a conventional bipedal robot with an articulated joint biped. The results showed that the biped with a prismatic knee joint achieved greater stability.

In quadruped robot locomotion using parallel mechanisms, researchers have used equal link lengths as legs for walking. However, force requirements are not the same in the forward and return strokes. An unsymmetrical parallel mechanism can be considered to accommodate such requirements. This work presents optimized dimensions of a 5R planar parallel mechanism (5R-PPM) with two degrees of freedom (DoF). Optimized dimensions are determined by formulating an optimization problem using kinematics and dynamics equations for the 5R-PPM. Genetic algorithm is considered to obtain solutions for the optimization problem formulated in this study. The constraint condition expressed here for optimization will attempt to minimize the peak torque essential to displace the links in the mechanism for the given height of the robot body and the path to be traced by the end-effector. After analysing all the four possible working modes for the same end-effector movement, the best working mode is selected for the quadruped legs. The equations are formulated and solved in MATLAB, and validated in the MATLAB Simscape Multibody toolbox.

Underactuated systems occur frequently in robotics and legged locomotion. Unactuated pendulum on an actuated cart is a classic example used for designing and testing control algorithms for underactuated systems. While pendulum balancing on a horizontally moving cart is popular and environments available for reinforcement learning, pendulum on vertically moving cart is rarely discussed due to relatively higher difficulty level in balancing it. This paper presents a model environment for a pendulum on a vertically moving cart and trains a neural network controller using reinforcement learning to balance it in vertical position indefinitely without exceeding the displacement limits. Results presented for both continuous and discrete force control input for the cart system show that the neural network controllers can successfully swing up and balance the pendulum.

A biped dynamic walker with two legs and two feet capable of walking in double support phase and allowing starting the gait cycle from rest position is proposed. The walker is actuated at the ankle joints alone with no hip actuator, fully actuated in double support phase, and unactuated in single support phase. Assuming a static configuration at the start of each single support phase, fixed point information for the gait cycle at various step lengths is extracted and represented with four parameters of a cubic polynomial. This is used as the end configuration for the position controller’s reference target in the double support phase. Actuation at the ankles considers the unilateral constraints at the front and rear feet. Even with trajectory tracking controller, low cost of transport is achieved by ensuring no negative power inputs during actuation. A proportional feedback controller is employed for cycle convergence, and the stability of gait cycles, disturbance handling, and energetic efficiency for various step lengths is shown through simulations.

This paper presents a robust biped dynamic walker based on the virtual slope, which introduces the driving force in the walking direction. Virtual passive dynamic walker as a reference model is used to develop the feedback controller based on the relation between the step length and the virtual slope of the walker. A numerical solution is prescribed to obtain the local and global stability of the biped dynamic walker. The Poincare map and the basin of attraction plots have been used for stability analysis. A fifth-order curve fitting polynomial function is used to set the relation between step length and slope. A proportional feedback controller is used to correct the error at the transition state.

Inverted pendulum control finds similarities with control of legged robots such as bipedal or humanoid robots where the trunk is balanced in an upright position. This paper proposes and presents a two degree of freedom cart that can move in the vertical plane while holding an underactuated pendulum. A neural network controller is trained using reinforcement learning to swing up and balance the pendulum by applying appropriate forces along horizontal and vertical directions. Simulation results show successful swinging up and balancing of the underactuated pendulum by applying horizontal and vertical forces on the cart while simultaneously keeping the cart within horizontal and vertical limits on displacements.

The biped dynamic walker considered in this paper has three actuators - two at the ankle joints and one at the hip joint. We consider the case of one of the two ankle actuators at fault. Despite having only two actuators operational, we show that successful gait is possible for a typical case of virtual passive dynamic walking. We analyze such gaits for local and global stability for a virtual slope and for the cases of completely unpowered or partially powered alternate steps. It is shown that completely unpowered alternate steps are preferred over partially powered alternate steps in the case of virtual passive dynamic walking for global stability, and the other way for local stability.

This paper presents the design of optimal dimensions for a two degrees of freedom parallel mechanism used in quadruped for walking application. Serial linkages or open link mechanisms have less stiffness and poor dynamic performance, thus parallel mechanisms were developed. Many researchers have used symmetrical parallel leg for quadruped walking but force requirements are different in forward and return stroke, thus unsymmetrical parallel leg may be optimal. Using genetic algorithm, optimum link length values are obtained and the corresponding peak torque is also found.

Legged locomotion is preferred over the wheeled locomotion as it can be used both for flat and rough terrains. Quadruped robots are preferred since they can offer better stability with considerable reliability. In recent years, passive dynamics has been used to obtain near zero-energy bounding gaits. Although theoretically such gaits consume no energy, in practice some additional energy is required to overcome losses. Existence and stability of such gaits have been thoroughly studied in literature for quadruped models with the assumption that the mass distribution and stiffness in the front and back legs are symmetric. Fixed points found using Poincare map indicate touchdown angle-liftoff angle symmetry between front and back legs. This property can be used to search for fixed points with ease. However, the range of initial conditions where the bounding gait is stable is highly limited. Control laws based on symmetry conditions observed are proposed in this paper to improve the stability region. One such control law based on body-fixed touchdown angles theoretically allows redesign of quadruped robot with physical cross coupling between legs to achieve inherent stability without leg actuation.

Robotic manipulators play an important role in industry for tracking as well as pick and place applications. Manipulators are required to carry out the commanded task with good repeatability and precision. With the help of robotic manipulators, very high speeds of operation are possible thus increasing the production rate. In order to achieve good position accuracy and reduce the vibrations of the end effector, these manipulators are made to have high stiffness of links. High stiffness designs make the manipulators heavy and bulky. Therefore, the existing heavy Rigid Link Manipulators (RLM) are shown to be inefficient in terms of power consumption or speed of operation with respect to the operating payload. Also, the operation of high precision robots is limited by the tip positioning accuracy requirement. Space manipulators due to their long arms have difficulties in controlling the tip positions and therefore have to work at sufficiently low speeds in order to reduce tip positioning error. Flexible link manipulators (FLM), which are much lighter and highly flexible compared to RLM, have been proposed in the past as means to reduce energy consumption and increase the speed of operation. Unlike RLM, the FLM has infinite degrees of freedom. Due to the distributed flexibility affecting the precision of operation, special control algorithms are required to make them usable. But FLMs due to their distributed flexibility have challenges in overcoming tip deflections for larger payloads and higher speeds of operation. Structural failures and larger tip errors for sufficiently large payload and high speed of operation limit the usage of FLMs. Though several methods to control tip error were studied by researchers but FLM could not make a place in industry today. In this thesis, FLM was modeled using finite segment approach due to its easy implementation and simulations were carried out to study tip oscillations for various payloads and speeds of operations. Position as well as torque trajectories were given as inputs to the hub and tip position was evaluated. Hub torque requirement for given trajectory was studied for FLM and RLM and the comparison is made. Study was also made to show that if FLM thickness is increase to tolerate higher stresses, an equivalent RLM can be constructed for the same mass having better tip position accuracy. Concept of dynamic workspace was introduced and it was shown that a single link FLM can be commanded at the hub to avoid obstacle which otherwise is not possible with single link RLM. A method to stiffen FLMs using thin cables, without adding significant inertia or adversely affecting the advantages of FLMs, has been proposed as a possible solution. FLM stiffened using thin cables can use existing control algorithms designed for RLMs. Cable stiffened flexible link manipulator (CSFLM) has been studied in detail in terms of the extent of stiffening and acceptable tip position error. Optimal string attachment locations was studied and results validated with the help of experiments. Spatial CSFLM was modeled using finite segment approach and simulation studies were presented. Limitations of CSFLM were discussed along with the methods to overcome. The CSFLM was shown as the potential candidate to replace the RLM in the industry and space applications saving manipulator cost, energy consumption, etc. Such cable stiffened flexible link manipulators have the advantage of allowing the use of existing control techniques designed for RLMs. Effectiveness of this new approach is shown through simulations and experiments.

Quadruped robots are sophisticated machines. They are equipped with several sensors in order to properly scan the terrain and look for possible positions where the foot can be placed. Here we have developed an electrically actuated quadruped robot to test the feasibility of such a robot. Generally most quadrupeds have pneumatic or hydraulic actuators which have greater force outputs but lack high accuracy. Electric actuators have higher accuracy but not very high force to weight ratio, hence the feasibility of an electrically actuated quadruped is explored.