Introduction: The BMX start is crucial for race performance, often measured by the time to the kink at 3.15 m from the start.
Objective: This study aims to optimize the BMX SX gate start using predictive optimal control techniques, focusing on the effects of maximal crank torque and reaction time on performance.
Method: Two models were used: The ‘upper extremities’ model analysed varying crank torques (250 Nm to 350 Nm) and reaction times (0.14 s, 0.16 s, 0.18 s).
The ‘two legs’ model was assessed under a single condition to better reproduce crank torque and track experimentally measured kinematics.
Results: Higher crank torques led to more forward initial positions, reduced recoil, and increased final velocities. The velocity of the start gate was a limiting factor initially, with timing and technique being crucial until the gate is halfway open. Reaction time variations showed minor effects on performance, and no strategy adaptation was needed within the tested range. The ‘two legs’ model accurately tracked experimental kinematics with low RMSE values. The predictive simulation with the ‘two legs’ model showed an improvement in kink time. The kink time for the predictive optimal control solution was 1.15 s compared to 1.23 s in the experimental trial.
Conclusion: This framework for researching the BMX start using predictive optimal control offers a systematic basis for future research. Insights can improve training strategies focusing on technique, timing, and initial start position. Future research could explore the effects of leg strength, hip range of motion, and gear ratios or crank lengths on performance.

The Dutch speedskating federation expressed the need for a feedback system for elite long-track speedskaters that can aid them in finding the optimal technique for an individual athlete. To contribute to this goal, this research aims to develop an optimization workflow that can reproduce realistic steady state speedskating behaviour. This is done with use of the simple skater model (SSM) (Van Der Kruk, Veeger, van der Helm, & Schwab, 2017). The research consists of two phases. The first to verify if the optimization can produce realistic speedskating motion, the second optimizes the speedskating technique to minimize the duration of one stroke. The optimization is solved with IPOPT. Even though the first phase can reproduce realistic speedskating motion, the optimal technique found in the second phase was unrealistic in terms of both trajectory and applied forces. This is caused by an inconsistency in the heading of the skate. This research shows the capabilities and limitations of optimizing the speedskating technique with the SSM.

In aerial manipulation, Unmanned Aerial Vehicles (UAVs) are equipped with manipulators to perform a variety of tasks such as inspections of critical infrastructure at heights. A fundamental issue is that the shaking forces and moments of the manipulator cause the UAV to tip-over and become unstable. Control based methods have been applied in which the UAV provided a compensation force or moment at the propellers. However, the dynamic model required was too complex to compute on-board in real time and simplifications led to poor performance. This thesis resolves the issue of shaking forces and moments by creating a new manipulator using inherent dynamic balancing principles. The advantage of these principles is that the manipulator architecture achieves both functions of supporting and positioning the end effector as well as balancing. This helps to reduce the weight of the manipulator. The result of the synthesis work is a manipulator which is reactionless, lightweight, has 3 degrees of freedom, and is compatible with a UAV. First a manipulator is designed using inherently force balanced architectures. Next, active moment balancing is developed through a novel control scheme. Finally, a simulation is performed to prove the dynamic balancing and control method. It shows the manipulator is reactionless. However, the control scheme’s tracking still needs improvement. This work is useful to enable UAVs with manipulators to perform a variety of tasks such as inspections of surfaces at height.

Haptic feedback from the steering wheel is one of the most important cues for driver to vehicle interaction. The right feedback is provided by ensuring that the haptic controller provides the required steering feel. Steering feel assessment and design is divided into a subjective and objective approach. The subjective approach entails experiments on the proving ground during which steering parameters can be tuned by steering experts. However, using only subjective assessment is time-consuming, costly and non-repetitive. Since there is no direct method to tune the steering feel objectively, a driver model is required to find a mathematical justification in the mechanical interaction between driver and vehicle during steering. A 3-dimensional multibody arm model is constructed to investigate the influence of driving posture on the nonlinear steering response. It was found that the torque acting in the shoulder joint is higher than in the elbow. The relation between joint torque and joint angles is
linear in the shoulder, whereas nonlinearities were found in the elbow joint. Nevertheless, a change of driving posture (i.e. a change of haptic interface) leads to a different steering response. Findings from the driver model were validated by two steering experiments. Muscle contraction was measured in order to analyse the forces acting on the joints.

This study shows promise to lead to a different approach for tuning steering parameters. Further investigation and detailed experiments are required to convert this driver model into a method to tune steering feel objectively.

In robotics, machine elements are accelerated in order for the machine to perform certain tasks, such as picking and placing objects. These accelerations result in inertia forces and inertia torques on the machine elements and on the base of the machine. These reaction forces and reaction torques on the base are called shaking forces and shaking moments. Shaking forces and shaking moments result in noise, vibration, wear and fatigue problems. Dynamic balancing eliminates shaking forces and shaking moments on the base, which results in low cycle times and high accuracy. However, the process of balancing a mechanism generally increases the masses, the moments of inertia, and the complexity of the mechanism.

The method of inherent dynamic balance aims at minimizing these drawbacks by considering the balancing prior to the kinematic synthesis. With the method of principal vectors, a large number of inherently shaking force balanced mechanisms has been found. However, the options for shaking moment balanced mechanisms are still limited.

In this thesis, an overview of current dynamic balancing methods is presented, along with a new method for the synthesis of inherently moment balanced mechanisms. This new method is used for the synthesis of inherently dynamically balanced 1-DoF pantographic linkages, where the desired motion of the end-effector is selected by the designer. This motion is defined as a set of precision positions. For this new method, the known method for RR chain synthesis from Burmester’s theory is combined with the shaking moment balancing condition. For the special case where the relationship between link angular velocities is linear, the shaking moment balancing condition is substituted into the RR chain design equation. For the general case where the relation between link angular velocities is non-linear, the equation of motion is numerically solved for a range of possible solutions in order to find the solution which reproduces the desired motion.

Current multi-DoF compliant manipulators are still rarely implemented in the industry because their range of motion (ROM) is limited as their designs are heavy and bulky or obtained by a serial set of multiple stacked flexure systems, which limits their compactness. The goal of this article is to overcome these limitations by considering them as an integrated multi-DoF compliant joint, either serial or parallel, and setting up a new method the Compliant Manipulator Design (COMAD)-method and investigate its performance. This method will combine the "Type synthesis of legs"-technique to include parallel kinematic solutions for the desired motion pattern whereafter the complete compliant solution space is obtained using the FACT-method. The method is applied for designing a 4-DoF-manipulator with a TTTR-motion pattern resulting in four new concepts composed of compactly aggregated wire flexures. After the concept selection, a demonstrator is manufactured which excellently possesses four decoupled motions with a relatively large ROM. This can be seen as a new milestone for designing multi-DoF compliant manipulators as it permits a larger ROM and better stiffness capabilities than those obtained from conventional methods because all compliant topologies are deflecting in series due to the parallel kinematic couplings within the multi-DoF flexure systems.

Background: The efficacy of a thrombectomy procedure for acute ischaemic stroke (AIS) is largely dependent on mechanical behaviour of thrombi and interactions with the thrombectomy device. Studies have examined clot analogues and evidence suggests that thrombus mechanical properties largely depend on composition, yet evidence based on thrombi retrieved from AIS patients is still lacking. Therefore, this study aims to characterize the mechanical properties of thrombi retrieved from AIS patients and to determine the relation to thrombus composition. Additionally, results were compared to literature in order to assess clot analogue representativeness.Methods: Directly following a thrombectomy procedure, unconfined compression tests were performed on thrombi retrieved from acute ischaemic stroke patients. For all tested samples, the material properties were characterized and related to the histologically determined composition. Identified histological components were 1) Fibrin & platelets, 2) Red blood cell and 3) Leukocytes. A subgroup analysis was performed to compare values with literature, where samples were stratified into four groups based on fibrin & platelet content (F&P -low, -moderate low, -moderate high and -high). Results: A total of 18 patients yielded 39 samples which were successfully tested and histologically analysed. Sample stiffness was found to be positively correlated to fibrin & platelet content (Rs=0.69, p<0.001). A good histological distribution was present within the data, as the fibrin & platelet content ranged from 7% to 99%. Subgroup analysis showed little difference in mechanical behaviour between the F&P moderate-low and F&P moderate-high subgroups, with the F&P low and F&P high groups respectively exhibiting a decreased and increased stiffness. Comparing to current literature, the results demonstrated that analogues most accurately resemble thrombi with a low fibrin & platelet content. Furthermore, all samples displayed viscoelastic and non-linear stress-strain behaviour.Conclusion: It was found that composition is a strong influencing factor of thrombus mechanical properties. Both at high and low fibrin & platelet contents, the relation between composition and stiffness was strongest, while it was least pronounced at moderate fibrin & platelet contents (approximately 25%-75%).

Wheeled-legged (hybrid) robots have the potential for highly agile and versatile locomotion in any real-world application requiring rapid, long-distance mobility skills on challenging terrain. The ability to walk and drive simultaneously is an attractive feature of these hybrid systems, but is unexplored in literature. This thesis work presents an online trajectory optimization framework for high-dimensional wheeled-legged quadrupedal robots where the feet and base trajectories are generated in a model predictive control fashion for robustness against disturbances. Our feet optimization employs a unique parameterization that captures the velocity constraints of the wheels’ rolling and our base optimization uses a ZMP-based balance criterion. Our approach is verified on a torque-controlled quadrupedal robot with nonsteerable wheels. The robot performs hybrid locomotion with different gait sequences on flat and rough terrain. Moreover, our optimization framework generates base trajectories at a rate of about 100 Hz and feet trajectories at 1000 Hz or higher. In addition, we validated the robotic platform at the Defense Advanced Research Projects Agency (DARPA) Subterranean Challenge, where the robot rapidly maps, navigates, and explores dynamic underground environments.

Instrumented treadmills and perturbations of the treadmill are commonly used for gait analysis and can provide real time biomechanical information and feedback of gait patterns and abnormalities. Force plates in the treadmill are combined with motion capture data and fed into a musculoskeletal model. High accuracy of the force plates is needed to give reliable feedback for gait analysis. The accuracy still needs to be tested under dynamic conditions, with the belt running. Also, inertial and gravitational forces are measured during perturbations as a result of the rotation and translation of the platform in which the force plates are positioned. This results in an error in the forces and moments as measured by the force plates, which is added up to the forces exerted by a subject. Inertia compensation models have been developed and showed promising results but have not been validated extensively. This study aimed to optimize and validate an inertia compensation model for perturbations on instrumented treadmills and validate the force measurements under dynamic conditions. It was shown that the treadmill can accurately measure the center of pressure (error = 1-6 mm), forces (error = 1-7 N) and moments (error = 0.5-4 Nm). A new calibration trial was found with higher sway accelerations which improved the inertia compensation model and left residuals forces and moments below 2 N(m). Moreover, it showed that using this inertia compensation model for pitch and sway trials led to a reduction of the kinetic residuals of up to 96% and values close to baseline measurements.

Because of the increasing e-commerce volume, resulting in increasing demands on the speed of delivery, logistical processes become more and more automated. Order picking is one of the last tasks that is done by humans in warehouses, because humans are flexible with respect to the large variability and changeability in items. However, it is a labour intensive and monotonous task, resulting in fatigue and a shortage of order picking personnel. This motivates the development of automated pick-and-place systems.
One of the challenges for such systems is the heterogeneity of items. In warehouses there is a big diversity in items so the system has to be able to deal with all of them. Another challenge is dealing with items that are deformable. Current systems often make use of suction cups but integrating sensors that can be used to handle deformable items is hard. Fingered robotic grippers have more potential in grasping these kind of items, but grasping deformable items is one of the least addressed topics in robotics. Therefore, the objective of this thesis is to design a control strategy for a fingered robotic gripper to grasp and hold deformable items in a pick-and-place task.
Inspired by the underlying principles that humans use to execute a pick-and-place task, a multi-level controller is proposed for a three-fingered gripper with capacitive pressure pads. The multi-level controller consists of a low-level computed torque controller and a high-level numerical optimisation based extremum seeking controller. The computed torque controller uses an internal model of the kinematics and dynamics, which is derived with screw theory, to compute the torques required to comply with the fundamental grasping constraint and the setpoint on the gripping force. The controller is tuned in such a way that the grasp quality is maximised, given a constant reference gripping force. Because of the fact that the properties of the items are unknown, an intelligent control system has to be able to determine the gripping force setpoint autonomously. This is the task of the high-level controller, that uses tactile sensors to derive the slip. This slip is used to determine the setpoint on the gripping force that the low-level controller has to follow, while maximising the grasp quality and not damaging the products as a result of applying excessive gripping force.
The proposed control strategy is tested and tuned in a simulation environment. The pick-and-place task is executed for the products from a virtual product inventory. The controller is optimised with respect to the control goal on a wide variety of deformable items. Designing controllers according to the proposed principle will increase the diversity of items that can be handled in a pick-and-place environment, while increasing the quality of the grasp and minimising the risk of damaged products.

An Oldham coupling is a constant velocity rotational transmission between two misaligned but parallel axis. Conventionally, this coupling uses two sliding contact prismatic joints, in which friction, wear, backlash and lubrication is unavoidable. Conversion to a compliant mechanism, in which motion is accomplished by elastic deformation instead, would eliminate all these drawbacks. Within PME, such a coupling is designed already. However, it was observed that it does not perform well for higher velocities and accelerations. Furthermore, it was indicated that little research is available about the dynamic behaviour of compliant mechanisms at all.
In this thesis, the dynamic performance of the family of compliant Oldham couplings is analysed and predicted. A straightforward generic analysis method is proposed, based on multibody dynamics. A case study is performed on the existing compliant design to validate the proposed modelling techniques. Its dynamic performance is evaluated experimentally and its failure mechanisms are indicated. Based on the gained knowledge, design improvements are proposed. Finally, this work now facilitates the design and implementation of compliant Oldham couplings in dynamic applications.

The continuous pursuit of reducing drag in many speed sports, such as cycling and ice-skating, demands novel approaches to gain further insight into flow phenomena around athletes. In recent years, the Ring of Fire measurement technique has emerged as a feasible option to visualise and analyse flow structures of transiting objects based on particle image velocimetry. The accuracy of this on-site measurement technique has not yet been validated under equal test conditions.

This master's thesis aims to compare drag area values of a cyclist from Ring of Fire measurements to simultaneously acquired power meter data. Tests with the cyclist in upright and time-trial posture, as well as different helmet types and various drafting distances, are envisaged to assess the correlation between the two measurement techniques in multiple drag area regimes, and to gain insight into large distance drafting above 3m, which, to the best knowledge of the author, has not yet been studied in academic research. In addition, the campaign plans to remove any user operations during the test, which would be another step towards a fully autonomous Ring of Fire system, as envisioned in the future. A spacious indoor facility is suggested as the testing site to minimize environmental effects and to allow for the continuous motion of the cyclist. The Ring of Fire method shows great potential, as the measurements are conducted under simulated racing conditions and wake visualisation allows the operator to locate origins of drag. Validating the drag area results could further attest to the Ring of Fire's viability as an optimisation tool in the upcoming years.

The conducted campaign, within the framework of this thesis, indicates good agreement between the power meter and Ring of Fire techniques when assessing the relative drag area delta of a small-scale helmet change and a large-scale posture change. In terms of absolute values, the power meter model shows a high dependency on underlying model constants. Using literature-based coefficients, the absolute CdA values are within 5% of the Ring of Fire derived values.

Furthermore, the feasibility of evaluating long distance drafting effects with the Ring of Fire system is demonstrated. Measurable drag area savings of 15% are obtained by the trailing cyclist at front wheel to front wheel distances of 7-9m. Due to non-uniform inflow conditions in front of the trailing cyclist, a wake contouring algorithm needed to be employed to satisfy mass preservation within the control volume by resizing the inlet and outlet plane. In addition, enclosure of the wake structure and contouring of a representative inlet plane is achieved. The flow topology in the wake of the trailing cyclist is acquired by a stereo-PIV system. Primary wake structures, as well as in-plane velocity fields, are comparable to those observed behind isolated riders.

The automotive industry is continuously improving the safety and comfort of the vehicles. To improve the safety and comfort of the vehicle more focus is put on improving driver assistance systems and also on making the vehicles more autonomous. In order for the safety systems to work properly and also to drive autonomously, it becomes essential that the vehicle has accurate knowledge of all the states of the vehicle. Two of such states are the roll and pitch angle. They are used to correct the images of a stereo camera, in roll control systems and for comfort assessment in autonomous vehicles.

Measuring the roll and pitch angle is in practice problematic. Gyroscopes are mostly used to calculate the roll and pitch angle by integrating the angular rates with respect to time. If low quality gyroscopes are used, this leads to inaccurate calculation of the roll and pitch angle, because the measurements contain noise or bias accumulated during the integration process. To use these low quality gyroscopes, they need to be incorporated in an robust algorithm together with other sensors. Only then the roll and pitch angle can be estimated accurately.

In this thesis a dynamic observer is developed, which has an internal vehicle dynamic model that calculates the tyre forces using an exponential tyre model. The performance of the observer is not only assessed using the true roll and pitch angle, but also compared against the performance of a kinematic observer.

To test the performance of both observers various test manoeuvres were executed in IPG CarMaker, which is a highly advanced vehicle simulation environment. The manoeuvres consisted of rapid steering inputs and hard braking to execute large roll and pitch angles. Sensor noise was added to the inputs of the observers to simulate low quality sensors. In this thesis it has been shown that incorporating dynamics into an observer significantly increases its estimation performance and outperforms the kinematic observer during all manoeuvres when sensor noise was added to the inputs of the observers.

The contribution of this thesis to the scientific field is that an exponential tyre model is used for the calculation of the roll and pitch angle in the internal vehicle dynamic model of the dynamic observer. Another contribution is that this thesis not only compares the performance of the dynamic observer against the real roll and pitch angles, but also against a kinematic observer, to see what the estimation improvement is when including dynamic relations.

The number of road accidents is increasing all over the world. When Vulnerable Road Users (VRU) are involved in an accident, they are prone to more serious injuries. Half of the fatalities around the world involve VRUs. One way of mitigating the severity and the number of VRU accidents is the introduction of Advanced Driver Assistance Systems (ADAS) functionalities. Such systems must be validated before being introduced on the market. Real-life tests for validating the system would imply driving around for millions of operational hours which would be very time consuming and therefore expensive. For this purpose, TNO has developed a scenario generation method and is currently extending it. The objective of this study was to develop a vehicle-cyclist interaction model which would then be used along the existing scenario generation method or for validating generated scenarios. The preceding literature study to this Master Thesis, from which the architecture of the models was developed and potential input parameters were defined, was used as a starting point. Based on the architecture, the interaction model was first developed with the help of a Hidden Markov Model (HMM). First, a sensitivity analysis was performed from which it was concluded that speed, acceleration, and steering angle are important parameters for a cyclist, whereas speed acceleration and yaw rate are important for a vehicle’s intent recognition. Furthermore, this analysis also indicated that distance to the center of the crossing is a very good interaction parameter. After selection of the parameters, the model was trained and cross-validated using data earlier recorded at TNO. The cross-validation results of the interaction model showed that the intent recognition is more constant for the whole observation sequence and an improved performance is seen for a prediction time of more than 2s compared to a HMM without interaction parameters. From this study, the kinematic input parameters for the cyclist and the vehicle, as well as the interaction parameters, were defined that need to be included in the vehicle-cyclist interaction model. However, the model developed in this study needs to be further validated using naturalistic vehicle-cyclist kinematic data, and for all type of vehicle-cyclist scenarios.

The pedestrian is regarded to be one of the most vulnerable road users. Non-verbal communication between drivers and pedestrians seems to play an important role in the mitigation of collisions. The emergence of autonomous vehicles in traffic in the near future presses the need to investigate objective measures related to pedestrian crossing behaviour and the efficacy of communication devices on autonomous vehicles that might replace the nonverbal signals of the human driver. In order to objectively investigate the efficacy of communication devices on autonomous vehicles, 24 participants in this study were immersed in a virtual reality environment, via the use of an Oculus Rift and an Xsens Link motion tracking device. In this virtual reality environment, participants were presented with 18 series of autonomous vehicles. Each series represented one unique combination of independent variables and contained a total of five vehicles. The vehicles were either equipped with a Text display or Frontal Braking Lights that indicated the yielding intentions of the vehicle, or were without any external interface. Furthermore, the inter-vehicular distance between the second and the third vehicle in the series varied between 20, 30 or 40 meters. The participants were instructed to cross the road onto the zebra crossing in the virtual environment when they deemed it was safe to do so. The experiment was designed in such a way that the only crossing opportunity for the participants was between the second and third vehicle when the third vehicles yielded. The road crossing decision of the participants, operationalized by the objective measure of their forward gait velocity, was earlier in time when there was either a Text display or Frontal Braking Lights present on the third vehicle in the series, when the inter-vehicular distance between the second and third vehicle was 20 meters and the third and subsequent vehicles yielded. Congruently, the self-reported ability of participants to predict the behaviour of the oncoming vehicles was significantly better when the third vehicle had a Text display compared to when there was no external interface. However, no significant difference in self-reported ability to predict the behaviour of the oncoming vehicles was found for the Frontal Braking Lights. Furthermore, the forward gait velocity was significantly greater in the presence of a Text display compared to when there was no external interface present for the condition in which the inter-vehicular distance between the second and third vehicle was 30 meters and the third and subsequent vehicles yielded. This work shows that besides the current standard of subjective validation by pedestrians of external human-machine interfaces on autonomous vehicles these interfaces can objectively be validated through the recording and differentiation of body motions.