Hassen Nigatu - Research

Research

Research Interests:

Continuum parallel robots ; Kinematics , dynamics and screw algebra ; Mechanism theory and development ; Reconfigurable mechanisms and robots ; Manipulation theory and multi fingered hands Rehabilitation and medical robotics ; Robotized packaging and industrial robotic ; Physical based animation and machine learning ; Quantum computation for robotics

Quantum Machine Learning and Grover's Algorithm for Quantum Optimization of Robotic Manipulators

A B S T R A C T

Optimizing high-degree-of-freedom robotic manipulators requires searching complex, high-dimensional configuration spaces, a task that is computationally challenging for classical methods. This letter introduces a quantum-native framework that integrates Quantum Machine Learning (QML) with Grover's algorithm to solve kinematic optimization problems efficiently. A parameterized quantum circuit is trained to approximate the forward kinematics model, which then constructs an oracle to identify optimal configurations. Grover's algorithm leverages this oracle to provide a quadratic reduction in search complexity. Demonstrated on simulated 1-DoF, 2-DoF, and dual-arm manipulator tasks, the method achieves significant speedups—up to 93x over classical optimizers like Nelder-Mead—as problem dimensionality increases. This work establishes a foundational, quantum-native framework for robot kinematic optimization, effectively bridging quantum computing and robotics problems.

A modified Lie group-based synthesis for parallel mechanisms: Addressing uncontrollable and immovable limb cases

A B S T R A C T

The synthesis of parallel manipulators using Lie group theory represents kinematic pairs through displacement operators, allowing the deduction of displacement subgroups from velocity fields using Lie-algebraic structures. However, two key limitations exist: (1) the effect of uncontrol- lable passive degrees of freedom, which are not reflected in the mechanism’s true degrees of freedom, and (2) the impact of ineffective joints and immovable limbs, which hinders the effective application of existing Lie group-based synthesis methods. These oversights limit the applicability of Lie group-based synthesis for certain parallel manipulators, leading to incorrect results during the intersection operation in synthesis. Failing to detect this phenomenon at the synthesis stage may lead to wrong mechanism. This paper introduces an enhanced method that integrates passive degrees of freedom and the effects of immovable limbs into the synthesis process, providing a correct representation of the motion in previously overlooked mechanisms. This approach extends the applicability of Lie group-based synthesis to various manipulators, ensuring consistent results, including for kinematically redundant parallel manipulators with passive

A Parallel Robot for Dynamic Balancing

The video shows a 3-limb R–R–S parallel manipulator performing high-precision tip–tilt balancing using a constraint-consistent 6×6 Jacobian and passive joint angle estimation, instead of direct platform pose sensing.

#Mechanism & #Hardware

Structure: Three identical R–R–S limbs, 120° apart on a fixed base.
Function: Controlled Rx–Ry tip–tilt (1T2R) with structurally constrained Tx, Ty, Rz; dependent (parasitic) motions are explicitly modeled.
Actuation: Base revolute joints driven by stepper/servo motors.
Sensing: Motor positions at actuated joints, embedded linear/matrix distance sensors for triangulated ball position, and base-mounted vision for monitoring and calibration.

#Kinematics

Full closed-loop model including actuated and passive joints.
Passive joint angles are estimated from geometric constraints to construct the digital twin.

#Control & #Balancing

Joint layer: High-rate position control with jerk-limited trajectories.
Kinematic layer: Uses the 6×6 Jacobian and constraints for synchronized joint commands.
Balancing layer: Uses ball position feedback to adapt tip–tilt and stabilize around the target.