Joint video-action models learn a shared representation to predict future video frames and robot actions, with decoupled heads that allow the video head to be dropped during inference for real-time policy execution. However, this decoupling means the two heads are trained independently against ground truth, with no mechanism for one head’s predictions to inform the other. We observe that the video head produces structurally incorrect robot poses, especially during dynamic motion. We introduce Kinematic Forcing (KF), a simple training objective that grounds this shared representation in the robot’s physical structure, using the action head’s own predicted pose as the source of that structure. From the action head’s predicted pose, we recover joint angles via inverse kinematics, render the robot with its URDF model, and composite it into the scene with an inpainter. We encode this kinematic composite using a frozen DINOv2 encoder and force the backbone’s shared representation to align with the features at robot-token positions using a physics-grounded, action-derived target. The entire rendering and alignment pipeline is training-only and removed at inference, leaving the deployed model identical to the baseline. Across simulated (WidowX, Google Robot) and real-world (xArm7) tasks, KF improves video prediction quality at robot regions, with the largest gains under fast motion, while maintaining downstream policy success.
... Read more

Diffusion Policies enable imitation learning by generating action sequences through iterative denoising. In robotics, policies are commonly trained centrally by pooling demonstrations into a single dataset, but demonstrations are often distributed across robots or sites and cannot be easily shared due to privacy, ownership, bandwidth, or operational constraints. This thesis studies whether diffusion based navigation policies can be trained with federated learning while maintaining closed loop performance under client heterogeneity and limited local data. We introduce FedDiff, a pipeline combining Diffusion Policy training with Federated Averaging (FedAvg), and evaluate it on a controlled 2D navigation benchmark with low dimensional beam observations.We compare three training regimes: (i) individual (no sharing), where each client trains only on its own data; (ii) centralized training on pooled data; and (iii) federated training via FedAvg. Experiments cover three scenario families: near-IID (clients have similar data distributions) with sufficient data, near-IID with scarce data, and non-IID compositional generalization across distinct navigation primitives. Across two sensing configurations (4-beam and 8-beam), we train 36 models and report success rates on held out start--goal configurations and, where applicable, on environments not used during training. Federated training is most beneficial under data scarcity and, in the 8-beam Near-IID setting, can match or slightly exceed centralized training on average, while single client training remains strongest on its own training room in the 4-beam data sufficient case. In the Non-IID setting, federated and centralized models perform strongly on primitive navigation skills but remain limited on composed environments that require combining those primitives. We also discuss training dynamics and practical sensitivities relevant to real world deployment. ... Read more

Fully actuated robots may be controlled using well-understood techniques, such as computed torque control, which override natural system dynamics. For underactuated robots these dynamics cannot be fully cancelled out, and must instead be leveraged, complicating the control problem. We present a classification of underactuated robotic systems based on the degree to which their dynamics can be decoupled. Finding coordinates that decouple the system dynamics simplifies control for two classes of robots, which we identify as partially- and fully decouplable robotic systems. In these decoupling coordinates, the Euler-Lagrange system representation has a block-diagonal inertia matrix and decoupled input matrix. After delving into the fundamentals of this proposed classification, this work implements an autoencoder as a first ML-based framework to learn these decoupling coordinates for the 2 degrees of freedom (DOF) case. Furthermore, we demonstrate how such representations simplify control. Input decoupling allows for collocated control using a straightforward PD + gravity compensation controller. Inertial decoupling enables non-collocated control through feedback linearization within a small set of states. To demonstrate the theory, decoupling coordinates are learned for a 2-DOF toy system. Performance of the learned coordinate transform is analysed, and controllers on learned and analytic decoupling coordinates are compared. ... Read more

This study presents the development and evaluation of an automated laser-based measurement system for determining the frequencies of Flexous oscillators while still in the wafer, using a Prusa I3 MK3 3D printer as a motion platform. Emphasis is placed on assessing the system's precision, repeatability, and edge detection accuracy. A lead screw-driven positioning mechanism with aluminium frame construction was designed to ensure high mechanical stability. Factors affecting positional accuracy, including microstepping, backlash, and structural deflection, are analysed and addressed. Calibration and homing tests confirmed sub-micrometre repeatability, with the Y-axis showing superior consistency. Movement accuracy tests revealed axis-dependent errors, influenced by mechanical and sensor-related factors. Edge detection performance was found to be sensitive to calibration speed, with slower movements significantly improving accuracy. The study concludes with recommendations for enhancing measurement precision through closed-loop control and predictive stopping logic, demonstrating the system’s potential for scalable, high-precision oscillator testing. ... Read more

... Read more

The loss of an upper limb significantly impacts an individual's life, directly affecting their ability to perform activities of daily living. Prosthetic devices play a crucial role in aiding amputees' rehabilitation in society, primarily driven by advancements in externally powered prostheses. Although prosthetic devices on the market offer excellent functionality, they come with a high price tag and use complex control algorithms.
Over recent years, a noticeable shift towards simplifying prosthetic devices has emerged, often facilitated by soft robotic principles. For example, the adaptive synergy approach has led to devices that are highly adaptable to their environment with a reduced degree of actuation (DoA), thereby improving functionality at low complexity. This thesis explores a novel research direction by combining the concept of adaptive synergy actuation with additional, parallel actuation of individual fingers. The main goal of this research is to assess the viability of using such a parallel adaptive synergy actuation structure for prosthetic hands. We designed a prototype incorporating this actuation structure, with the main design goals of high functionality, low complexity, robustness, and anthropomorphic sizing. The hand, which features a tendon-driven design, has 15 joints, including 14 dislocatable joints and one revolute hinge joint. The entire hand is powered by a single primary actuator, with two smaller additional motors operating in parallel on the index and thumb. We empirically validated the prototype's performance through qualitative experiments, and its performance was compared to that of other prosthetic devices available on the market through quantitative analysis. Evaluation of the prototype revealed promising results, such as its ability to adaptively grasp various functional objects and execute complex tasks. Force measurements revealed performance comparable to devices on the market. The results indicate that this novel actuation principle, with future refinement, is an interesting new approach to increasing functionality with minimal increase in complexity and can offer an excellent alternative to costly prosthetic devices currently on the market, thereby enhancing the accessibility of functional prosthetic devices for individuals with upper limb loss and improving their quality of life. ... Read more

This work introduces a novel methodology for the development of interpretable reduced-order dynamic models specifically tailored for jumping quadruped robots. Leveraging Symbolic Regression combined with autoencoder neural networks, the framework autonomously derives symbolic equations from data and fundamental physics principles capturing the complex dynamics of jumping actions with high fidelity. This approach significantly reduces model complexity while enhancing interpretability, facilitating deeper insights for legged robotic applications. The efficacy and accuracy of the proposed models are validated through comprehensive experimental studies, marking a substantial advancement in the design of agile and efficient legged robots. This research demonstrates the outperformance of a learned 2D model compared to existing template models such as the ASLIP. Also, an analysis of the dimensionality of the learned model is conducted showing the overarching tradeoff between accuracy and complexity. The method is validated on different simulated quadrupeds and an actual hardware robot. ... Read more

In recent years, soft robots have become a focal point of research due to their ability to mimic natural movements and adapt to unstructured settings. However, their inherent flexibility poses significant challenges, particularly in the areas of modelling and control. While data-driven methods can model soft robot behavior without explicit physical models, they require extensive data and lack interpretability. On the other hand, physics-based low-dimensional models have relied heavily on expert knowledge and intuition, sometimes leading to models that are either too simple and inaccurate, or excessively high-dimensional.

This thesis introduces an end-to-end methodology for automatically identifying low-dimensional kinematic and dynamic models of planar continuum soft robots using image data. Based on the Piecewise Constant Strain (PCS) parametrization, the proposed approach determines an efficient segmentation for the soft robot to approximate its configuration. Afterward, a model identification strategy is employed to obtain a dynamic model that contains only the most essential strains. This model is formulated in the standard Euler-Lagrange framework, facilitating the integration with conventional model-based control schemes. The methodology is validated through simulations involving various planar soft manipulators and in the presence of noise, demonstrating its capability to generate accurate and computationally efficient models. This work provides a fast and practical tool to help the modelling and control of continuum soft robots, highlighting the potential for future applications in more complex actuation systems and real-world soft robots. ... Read more

Experience replay for off-policy reinforcement learning has been shown to improve sample efficiency and stabilize training. However, typical uniformly sampled replay includes many irrelevant samples for the agent to reach good performance. We introduce Action Sensitive Experience Replay (ASER), a method to prioritize samples in the replay buffer and selectively model parts of the state-space more accurately where choosing sub-optimal actions has a larger effect on the return. We experimentally show that this can make training more sample efficient and that this allows smaller function approximators -- like neural networks with few neurons -- to achieve good performance in environments where they would otherwise struggle. ... Read more

Soft robots have the potential to accelerate robotiza- tion in areas that are complex and impractical for hard robots. The use of soft materials results in a safe and flexible design that is unattainable for hard robots. However, this attribute results in the need for new control approaches and strategies. Hybrid controllers are a relative unexplored type of controllers that consist of a model-based controller part and a learning part to correct the model-based controller. A hybrid controller benefit by the unrequired need for accurate system identification. Simultaneously, the learning effort is reduced by the preliminary work of the model-based component. In this project, a model-based feedforward controller is pro- posed and compared with a hybrid controller consisting of the same model-based controller enhanced with a Gaussian process to reduce the end-point error in the bending angle. The controllers are tested using a crafted 2-segment pneumatic silicone soft robot, following a circular trajectory with different radii. The results of this new control strategy highlights the poten- tial benefits of adding a learning approach to a model-based controller to reduce model errors. Using a relative small dataset preserves a computational usable Gaussian process. The small dataset remains effective by reducing the range of the training data. ... Read more

This paper presents a novel approach to regional forecasting of SARS-Cov-2 infections one week ahead, which involves developing a municipality level COVID-19 dataset of the Netherlands and using a spatio-temporal graph neural network (GNN) to predict the number of infections. The developed model captures the spread of infectious diseases within municipalities over time using Gated Recurrent Units (GRUs) and the spatial interactions between municipalities using GATv2 layers. To the best of our knowledge, this model is the first to incorporate sewage data, the stringency index, and commuting information into GNN-based infection prediction.
In experiments on the developed real-world dataset, we demonstrate that the model outperforms simple baselines and purely spatial or temporal models for the COVID-19 wild type, alpha, and delta variants. In combination with an average R2 of 0.795 for forecasting infections and of 0.899 for predicting the associated trend of these variants, we conclude that the model is well suited for predicting the spread of infectious diseases with similar disease dynamics in real world applications. To increase prediction performance and to improve the generalizability of the model for infectious diseases with more complex disease
dynamics, we recommend using additional (synthetic) data or expanding the regional forecasting scale in future work. ... Read more

Reliable indoor navigation in the presence of dynamic obstacles is an essential capability for mobile robot deployment. Previous work on robot navigation focuses on expanding the network structure and hardware setup leading to more complex and costly systems. The accompanying physical demonstrations are often limited to slow-moving agents and simplistic obstacle configurations. In this thesis, we develop an end-to-end navigation system with a focus on real-world transferability and produce a low-cost and customizable robot platform. Instead of expanding the network structure, we rely on external capabilities such as backward motion, frame stacking, and behavioral reward design to improve performance while preserving transferability. By convention, these methods have been largely disregarded in previous works on deep reinforcement learning (DRL) for unmanned ground vehicle (UGV) navigation. We analyze the effect on performance in simulation of different off-policy algorithms with hyperparameter and reward function configurations. Experimental results show that our agent can achieve state-of-the-art performance in challenging and unseen simulated environments. In addition, physical robot demonstrations show that our system is capable of dealing with fast-moving and unpredictable agents in a real-world environment. ... Read more

The Free Energy Principle, which underlies Active Inference (AI), is a way to explain human perception and behaviour. Previous literature has hinted at a relation between AI and Linear-Quadratic Gaussian (LQG) control, the latter being a textbook controller. AI and LQG are, however, defined with different settings in mind: LQG has access to inputs, whereas AI estimates these; LQG is optimal for White Gaussian Noise, whereas noise needs to be coloured for AI, in order to make derivatives. Therefore, a comparison is provided on two bases: the setting of LQG control; and the setting of AI. The optimal LQG controller is obtained for both settings, and AI is applied to both settings as well. When AI is reduced to the setting of LQG, an equivalent expression can be obtained by a proper choice of tuning parameters. This entails choosing a matrix such that the closed-loop is stable, which contrasts LQG control, which is always a stabilizing controller. When LQG is extended closer to the normal AI setting, we find that tuning of AI becomes harder for more complex systems, but that AI is mostly able to show optimal behaviour. ... Read more

Robots that use legged locomotion have the ability to overcome obstacles and can negotiate a wide range of difficult terrains, such as encountered in outer-space missions. In many practical scenarios however, their applicability is still limited, mainly due to insufficient speed and efficiency. On the other hand, robots that use wheeled locomotion are fast and efficient, but are generally confined to flat or prepared surfaces. A relatively new approach, to use the advantages of both forms, is the combination of walking and driving technology into Hybrid Walker-Wheeler technology. In this research we will explore the feasibility of applying Hybrid Walker-Wheeler technology to the Zesbenige Robot (ZebRo). ZebRo is a small walking robot with six One-Degree-of-Freedom (DoF) legs and walks with an insect-inspired gait. It is being developed with the intention to go on a mission to the moon. The objective in this thesis is to increase the speed and energy efficiency of the ZebRo on flat surfaces, while maintaining its robustness and walking capability on rough terrain. We will set up the criteria for a new design, explore various options and parameters and choose a concept. We will then do a number of simulations to analyse the properties of a wheel and a leg and to apply these in the design. The final prototype consists of a single module with a wheel, a one-DoF leg and a custom-design coupling which switches torque, from the motor, between the wheel-axle and the leg-axle. This prototype was tested and evaluated on its electrical power consumption and the torque and speed transmitted to the leg-axle and wheel-axle. From these results, we were able to draw a number of conclusions and make recommendations for a ZebRo equipped with Hybrid Walker-Wheeler technology. ... Read more

Hand pose tracking in 3D configuration is one of the popular components of human-computer interaction research, which can be seen applying in virtual reality (VR) and augmented reality (AR). Although the research field has been thoroughly studied in the past few decades, most methods emphasized recognizing hand gestures or joint locations. However, computer interaction these days requires numerous manual object interactions. The paper first reviews state-of-the-art hand pose estimation and tracking methods to achieve accurate hand tracking and control. Then we describe the general procedure for recovery 3D hand model. In the second place, the results of tracking whole hand kinematics using the extended Kalman filter (EKF) based on data recorded from the surface markers are presented. Finally, in the part of teleoperation, the paper presents a novel method for applying synergy-based mapping strategy for SoftHand 2 control. And we demonstrate the simulation results of telemanipulation with hand racking through Simulink. ... Read more

The scheduling algorithm of the printer is an important factor that affects printing efficiency. For current printers, paper scheduling often follows the first-in-first-out principle, so it is often not optimal. The printer system is a type of semi-cyclic discrete-event system with synchronization but no concurrency. The system contains a set of operations that vary over a limited sequence of operations, which can be further modelled as a Switching Max-Plus-Linear (SMPL) system through max-plus-linear algebra, and can be switched between different modes of operation. Since max-plus algebra has a significant analogy with conventional algebra, some properties in conventional algebra can be applied to such kind of system, making the system easier to achieve optimal control.

In this report, a printing scheduler modelled by the max-plus algebra is presented. We review and summarize the basic knowledge of the Max-Plus-Linear (MPL) system from the existing literature in the first half of the report. We first introduce the basic properties of max-plus algebra and SMPL systems. Then we turn to the stochastic case, and consider the two types of stochastic uncertainty that may be included in the SMPL system, namely stochastic parametric uncertainty and stochastic mode switching uncertainty. Next, we review a Model Predictive Control (MPC) approach that can achieve optimal scheduling of systems containing two random uncertainties.

In the second half of the report, based on the content summarized in the first half, we make a further derivation of the printer scheduling system modelled by SMPL approach. We first present the modelling framework of the printer. Three working modes, namely duplex mode, idle mode and simplex mode, are modelled separately and merged into a compact form. Then a scheduler that takes the feeding and processing time of each sheet of paper as design variables is introduced. It can find the global optimal schedule of different types of paper by solving the Mixed-Integer Linear Programming (MILP) problem. After that, we discuss cases involving switching, and consider switching between two sizes of paper and two working modes. Three possible intermediate switching modes are designed. Finally, we take noise/interference into consideration, discuss the impact of noise on scheduling, and the changes that the previously proposed methods may require to achieve optimal scheduling in the presence of noise. ... Read more

A fundamental task of intelligent and autonomous robots is to infer from observations the state of the world. This inference is generally achieved by employing a filter, which consists of a model and filtering law. Learning this model and filtering law from observations is another fundamental part of robotics, and is generally referred to as system identification.

Neuroscientist K.J. Friston has developed a relatively novel theory on biologically plausible human brain inference called the Free-Energy Principle. One of the theories within the Free-Energy Principle, namely that Dynamic Expectation Maximization (DEM), has been suggested as a novel method for filtering and system identification. This method is expected to outperform standard Expectation Maximization (EM) in terms of hidden state and parameter estimation in settings where noise is correlated. However, in order for this neuroscientific theory to be properly used for robot inference, two problems must first be solved.

The first of these problems is the fact that the theory is defined in the continuous-time domain, whereas data available for system identification is always discrete. In this thesis I will suggest three discrete-time interpretations for DEM-based system identification. The major difference between the three methods is the information that is embedded in the generalized signals: predictions, derivatives and past data.

The second problem is that the filtering method corresponding to the Free-Energy Principle depends on data which is not available: the derivative signals of measured in- and outputs. I introduce two fundamentally different solutions to this feasibility issue: a numerical differentiator and a stable filter. Both of these solutions are shown to find an estimate for the unavailable data. However, the former is shown to significantly outperform the latter.

Furthermore, the theory described in this thesis is implemented into a novel python toolbox for system identification. This toolbox can be used as a basis for further research and be approved along with it, until at some point it is ready to be used for real applications.

Using the toolbox, the DEM-based identification and filtering methods are tested though various numerical simulations and the results are compared with the EM method. Results show that with the implemented settings none of the suggested discrete-time filtering methods outperforms the conventional Kalman filter. The main cause of this inferior performance is shown to be instability in the filtering method. I make some suggestions for overcoming this problem. As a result of the inferior performance, the joint-performance of the suggested DEM-based parameter- and state- estimation methods also proves to be inferior in terms of parameter estimation accuracy.

However, results show that the theoretical parameter optima of the Free-Energy as determined from known hidden states are in fact close on the real parameters, and furthermore show to be invariant to noise correlation. This suggests that should the instability issue as some point be solved and a better means to approximate the theoretical optimum be found, the DEM-based methods might in fact outperform EM both in terms of hidden-state and parameter accuracy in settings with correlated noise.
... Read more

A wet clutch is a device that transfers torque between two shafts via a hydraulic mechanism. Wet clutch control is key to achieve smooth and fast clutch engagements. Optimal control of a wet clutch is not trivial because of the complexity of the system due to nonlinearities, hybrid dynamics and changing dynamics over time due to changing temperatures and wear. Nowadays simple parametrized feedforward controllers are used in industry. The parameters of the control signal are tuned by hand and updated over time by an operator to account for the changing dynamics. Several solutions to the wet clutch control problem exist in the literature, however the structure of the solutions is fixed beforehand and they rely on expert knowledge of the system. Another challenging aspect of wet clutch control is model uncertainty, currently robustness of performance is considered by running a finite number of experiments, but no hard guarantees can be given. In this thesis a method is developed to automatically synthesize controllers for a wet clutch which are robust to model uncertainties. The method uses Genetic Programming (GP) to automatically synthesize controllers. Using GP controllers for a wet clutch can synthesize without fixing the structure beforehand and in a multi-objective way. Robustness is considered by optimizing the worst case performance. To be able to give a guarantee on the worst case performance of a controller a method, that is able to formally guarantee a lower bound on the worst case performance using reachability analysis, is developed. This method is to costly to incorporate into GP and instead a use cheap estimation of the worst performance in practice.
This method was able to find robust controllers which outperformed a hand tuned baseline controller, but in order to compare this method to other methods in literature, real life experiments are needed. ... Read more

Wireless power transfer is getting more attention than ever before but seems to be lagging in the go-kart industry. Manufacturers are designing more electric powered go-karts, but haven’t found a proper solution for most of its major drawbacks. The added weight from the battery pack and the required time to recharge are just two of these major drawbacks. These drawbacks make electric go-karts less desired in the rental industry, where turn-around-time is required to be as short as possible; in order to maximize profit. An Inductive Power Transfer System can be used to charge the battery pack of an electric go-kart statically in the pit-lane, but also dynamically on the track. Both of which are shown to reduce the required size of the battery-pack of electric cars and busses. This study determined the feasibility of an IPTS on go-karts, considering the current state of the art, in order to solve these drawbacks. The design of the charger was optimized for a go-kart while keeping its most important features the same. A safety analysis was performed using a FEM-program to see whether the system can be used by the public. The go-kart with the charger was modeled and simulations were used to determine the required size of the battery pack in a case-study. The results of this case-study were used to evaluate the financial viability of an IPTS on go-karts. ... Read more

Pioneering Spirit, Allseas's largest pipelay vessel, will be outfitted with a novel Jacket Lift System (JLS). A jacket refers to the steel frame which supports the topside of a fixed offshore platform. The Pioneering Spirit was already capable of lifting the topsides of offshore platforms, like the former oil platform Brent Delta, but would have to leave the jacket behind. With the JLS it will also be able to remove and install jackets.

In offshore ship mounted crane applications wave disturbances can create unwanted oscillations in suspended loads. Anti-sway control systems have been developed to aid the crane operator in decreasing these oscillations. But not all crane types have actuators suitable for this compensation. The JLS will be a derrick type crane which will be positioned above a jacket using the dynamic positioning of the ship. The hoist systems of the JLS are then lowered and connected to the jacket. During this lowering process the ship motion induced by wave disturbances create a large oscillation in the hoist system.

In this thesis a controller is developed based on a proportional derivative controller found in literature to reduce the sway in the hoist system to make the connection process possible. ... Read more