Remotely controlled vehicles have gained increased interest and application, like the exploration of inhospitable environments. To this end hexapods with C-shape locomotors are particularly suitable because these vehicles possess both the efficiency of wheels and the climbing capabilities of legged robots. Currently all C-legged hexapods are equipped with the same C-shape tip starting in the center of rotation, yet few to no alternative shapes that could improve the performance have been investigated. The performance is mainly characterized by three measures: traction, climbing and mobility. However, it was also found that no system level performance analysis including constraints exists. In this work a novel method is proposed to describe the system level performance of C-legged hexapods. Using this model, a computer-aided manual optimization of the leg shape is performed, from which it is observed that there is no unique optimal leg shape. The optima depend on the weight factor posed by the designer; but several leg types prove to perform better than the conventional shape. The most prominent trade-off of this hybrid shape is present between climbing and mobility, but this effect can be superseded when the shape is generated conceptually instead of analytically.

In this paper the first gripper that can grasp bunches of bananas is presented. Bunches of bananas are difficult to grasp by a mechanical gripper, as they vary in size and shape and are sensitive to damage caused by mechanical impact. In former research, single bananas have been grasped with pinching- or granular grippers. The gripper proposed in this paper grasps the bunch at its most sturdy part, namely its tip, and uses a hook finger to minimize the chances on damage. The scissoring fingers allow the gripper to handle the variability of the object. Additionally, the scissoring combined with the funnel shaped finger tips ensure that the gripper compensates the position errors of robotic systems. In the evaluation of its abilities the gripper is able to pick up 19 out of 19 bunches of bananas. It has a tolerance for a maximum positioning error of 4.9 cm and causes only minor damage on the peduncles of the bananas. In practical experiments, the gripper performed adequately and showed to be able to cooperate with the object to guide itself to the right location. The novel strategies of hook gripping and the extra degree of freedom of the compliant X-joint show that bunches of bananas can be grasped mechanically, while compensating for some of the shortcomings of current robotic systems.

In mechanisms and machines, elements' motions can generate reaction forces and moments on the base of the mechanism, which are called shaking forces and shaking moments. These induce vibrations of the base, which create noise, wear and fatigue problems and reduce the accuracy of the systems. Dynamic balancing is a solution to eliminate these reaction forces and moments by generally introducing additional counterweights and counter-rotating elements. Mechanisms having zero shaking forces and moments are called, respectively, force balanced and moment balanced. Dynamically balanced mechanisms are both force and moment balanced. Since the introduction of additional elements increases the total mass and inertia of the mechanisms, the method of inherent dynamic balancing aims at designing dynamically balanced mechanisms which do not include additional elements. By considering dynamic balance as a design principle, all the links contribute to both the motion and the balance of the mechanisms. These are called inherently dynamically balanced mechanisms and can be synthesized from inherently force balanced linkage architectures, which are based on principal vectors. However, the design of these architectures consists in parallelogram linkages which can potentially compromise the force balance when their links become collinear. In addition, links can overlap and represent a potential limitation in real applications. This thesis presents techniques which modify the linkage architectures and can prevent the potential issues related to their original design. The number of degrees of freedom can be reduced and specific motions can be created by constraining links’ rotations and translations. Moreover, parallelograms’ sizes can be modified and links can be replaced by machine elements like sliders, gears, belt and chain drives. It will be shown how force balance is maintained after having modified the linkage architectures.