Planetary exploration robot teams rely on robust exploration pipelines, such as multi-agent SLAM, to perform complex collaborative tasks. Multi-agent SLAM methods rely on accurate relative transformations between agents to collaboratively merge maps, predominantly using indirect associations between keyframes to compute transforms. Recently, direct association methods that use onboard perception modules with pose estimation pipelines have been shown to provide more accurate, long-range relative transformations. However, current methods consider the robot as a single static rigid body and struggle greatly during dynamic movements, complex robot morphologies, and under harsh environmental occlusions. As a potential solution, we represent the robot as an ensemble of articulated components and investigate the use of articulated pose estima tion methods with RGB images and intrinsic measurements. Specifically, we introduce novel post-processing steps based on factor graphs and differentiable rendering to refine upstream poses. First, pose estimates of individual components are visually refined using differentiable rendering and a segmentation mask to ensure visual plausibility. Second, by using a factor-graph formulation, we fuse estimated poses with partially available intrinsic measurements to preserve relative robot constraints and reconstruct missing estimates under occlusions. Third, we extend the factor-graph refinement to estimate over multiple frames and use visual odometry data (VO) to mitigate rotational symmetry issues. Experimental results on synthetic planetary terrains demonstrate that our method outperforms single-body pose estimation with a decrease in translational and rotational error by 55% and 61%, respectively. Additionally, our novel factor-graph refinement method is able to recover undetected articulated components under heavy occlusions and provides robust performance under dynamic movements in indoor lab data. ... Read more

Fully autonomous driving has long been promised, yet remains difficult to deploy at scale. Safe and efficient autonomy hinges on motion planning and control under tight constraints, realistic vehicle dynamics, and reliable cooperation between vehicles, often in the presence of delays, uncertainty, and adversarial communication. Progress in these areas depends on fast design–test cycles, but current validation workflows are polarized, high-fidelity simulation enables rapid iteration, yet struggles to reproduce road–tire interaction, actuator and sensor dynamics, and multi-agent network effects. Full-scale experiments capture these effects, but are costly, slow, and constrained by safety and regulatory requirements. This thesis addresses this gap by enabling physically grounded experimentation by designing a low cost and reproducible robotic platform, and by developing control and coordination methods for cooperative driving applications for more efficient traffic flow, as well as model based control strategies to handle safety critical maneuvers at the limits of handling.

Three limitations in the state of the art motivate the research objectives of this thesis. First, existing small-scale car-like platforms either remain relatively expensive for multi-agent experiments, depend on custom components that hinder reproducibility, or lack rigorous system-identification workflows needed for model-based control and sim-to-hardware transfer. Second, cooperative driving methods, for example platooning, often rely on timely attack detection or hard switching to degraded modes, leaving vulnerability windows and risking undesirable transients, moreover, adversarial robustness at the coordination and topology-management level, such as merge, split, rearrange, is less developed than low-level spacing control. Third, widely used MPCC formulations can become unreliable on highly curved paths because internal progress approximations drift under large deviations, which can degrade constraint handling, racing-oriented MPC variants frequently depend on offline priors, for example precomputed racelines, limiting transfer to emergency-like maneuvers where only local geometry is available. Accordingly, this thesis aims to develop: (I) a reproducible, low-cost experimental platform with a complete identification pipeline, (II) cooperative driving algorithms that preserve safety even under persistent or undetected communication attacks and remain feasible under actuator saturation, and (III) curvature-aware MPCC formulations that remain reliable on tight curvature and avoid dependence on offline priors in high-performance settings.

The thesis begins with the design and characterization of the Delft Autonomous-driving Robotic Testbed (DART). DART is a low-cost, reproducible, small-scale vehicle built on a commercial RC chassis and augmented with additional sensing and computational capabilities, including Lidar, an IMU, custom wheel encoders, and onboard computing. The platform preserves the essential features of full-scale vehicle dynamics, Ackermann steering, suspensions, electric motor, while remaining small enough to work with in laboratory settings. A central component is a comprehensive system identification procedure that yields reliable kinematic and vehicle dynamics models tailored to small-scale vehicles, along with sub-models for motor force, friction, steering actuation, and tyre lateral forces. This modelling pipeline enables realistic testing and accurate model-based control on hardware.

The thesis then shifts focus to cooperative multi-vehicle systems by introducing a distributed, attack-resilient platooning framework. This work addresses two intertwined challenges: maintaining safety and formation integrity under malicious interference on the communication channels, and enabling coordinated platoon-level decisions such as merging, splitting, and rearranging. At the control level, the method combines sensor-based Adaptive Cruise Control with communication-based Cooperative Adaptive Cruise Control, while a safety filter ensures collision avoidance even when communicated acceleration data is corrupted. At the coordination level, a distributed topology-management strategy detects inconsistent information and isolates compromised vehicles by reorganizing the platoon. The approach is validated in simulation and experimentally on multiple DART vehicles, showing that safety and string stability are preserved despite communication attacks.

With this platform at hand, we then focus on designing a motion-planning strategy for urban driving. We investigate model predictive control and develop a Curvature-Aware Model Predictive Contouring Control (CA-MPCC) framework. Traditional MPCC formulations assume low curvature and rely on a lag-error term that couples progression along the reference path with lateral tracking. This complicates tuning and reduces reliability in tight curves. The CA-MPCC formulation resolves these limitations by explicitly accounting for curvature in the path geometry, removing the lag-error term entirely and simplifying both the cost structure and the tuning process. The method is validated in simulation and on the DART platform, demonstrating improved robustness, reduced parameter sensitivity, and reliable real-time performance even on highly curved trajectories.

Next, the thesis extends the curvature-aware MPCC methodology to high-performance domains with the rCA-MPCC framework for autonomous racing. Racing introduces additional challenges, such as operating at the limits of handling, rapid curvature changes, and strong coupling between dynamical states. The rCA-MPCC formulation augments CA-MPCC with a curvature-informed terminal cost and a compact reference-path representation that avoids dependence on precomputed race line information. This chapter also expands the physical vehicle model through actuator-dynamics identification and residual dynamic modeling using Gaussian Processes. Together, these advances improve prediction accuracy and enable robust high-speed control. We furthermore extend the car-racing formulation to aerial drone racing, bridging the gap between the two communities. Experiments on small-scale cars and simulations on quadrotor drones demonstrate faster, more consistent lap times and robustness across different dynamic model choices.

Finally, the thesis demonstrates how the developed platform and control tools can be applied to cooperative multi-robot tasks such as persistent monitoring and target detection. The chapter combines DART with the CA-MPCC controller to implement Lissajous-curve-based coordinated trajectories generated by time-inverted Kuramoto dynamics. The result is a distributed coordination strategy that guarantees collision avoidance and complete area coverage. Experimental validation confirms that the high-level coordination method and low-level CA-MPCC controller integrate smoothly on real hardware, even under disturbances and temporary agent failures.

Overall, this thesis contributes a unified framework for physically grounded experimentation, accurate dynamics modeling, and robust control and coordination in autonomous driving. By developing a reproducible platform, constructing reliable models, and demonstrating advanced control strategies, from emergency maneuvers to cooperative platooning and multi-agent monitoring, it lowers the barrier to real-world validation and accelerates iteration cycles for autonomous driving research. The insights presented here support future work toward safer, more robust, and more efficient autonomous transportation systems. ... Read more

Autonomous mobile robots have become increasingly capable over the past decades, enabling their use in domains such as logistics, agriculture, and healthcare. This thesis contributes to a collaborative project between the Dutch National Police and the Delft University of Technology, focusing on search and rescue operations in the security domain.

In these safety-critical missions, robots are expected to support human first responders by exploring unknown and potentially hazardous environments, including buildings and natural terrains. To reduce the cognitive and operational burden on human operators, robots must navigate cluttered spaces both autonomously and safely. In the context of this thesis, autonomy refers to onboard decision-making and safety to avoiding collisions despite uncertainties like sensor noise or environmental effects.

This thesis explores autonomous and safe navigation using model predictive control (MPC), a planning and control strategy that predicts future behavior based on a system model and optimizes actions accordingly. MPC’s strength lies in its ability to handle constraints such as actuator limits and obstacle avoidance. However, applying MPC in real-world navigation presents several challenges, which this thesis addresses through the following contributions.

Navigating a mobile robot through cluttered environments requires it to follow dynamic trajectories while avoiding obstacles. Although MPC is well-suited for this task, recent methods for collision-free trajectory tracking often rely on complex mathematical formulations that can be difficult to interpret and apply. Chapter 2 aims to bridge this gap by offering a structured, step-by-step guide. It explains how to model the robot’s navigation problem, formulate the corresponding MPC problem, and establish performance and safety guarantees.

The chapter presents three formulations: a nominal MPC approach for ideal conditions, a robust MPC formulation that accounts for bounded disturbances, and a robust output-feedback MPC formulation that additionally handles measurement noise. Each formulation is supported by theoretical insights and practical considerations. While not exhaustive, the guide is intended to support researchers and practitioners in implementing MPC-based navigation under varying levels of uncertainty.

To enable autonomy, the MPC formulations introduced in Chapter 2 must operate in real time. This is challenging because MPC relies on solving an optimization problem at each time step, which can be computationally demanding. Planning long-term trajectories and computing control commands at high frequency on embedded hardware is especially difficult.

Chapter 3 addresses this by introducing a hierarchical MPC (HMPC) framework that separates planning and control into two layers. The planning MPC handles long-term trajectory generation at a lower frequency, while the tracking MPC focuses on short-term execution at a higher frequency. This separation allows the use of complex nonlinear models in both layers without compromising real-time performance. The tracking MPC layer, freed from long-term planning, can focus on precise tracking, improving stability and responsiveness.

The HMPC framework also includes a method for generating consistent collision avoidance constraints. Its effectiveness is demonstrated through simulations and lab experiments, showing safer and approximately four times faster goal-reaching compared to a single-layer MPC approach.

While the HMPC framework improves autonomy, it leverages a nominal MPC formulation that assumes perfect models and accurate state data. In practice, mobile robots operate in the presence of model uncertainties and noisy measurements, which can lead to constraint violations such as collisions.

Chapter 4 extends the HMPC framework to address these issues by incorporating robust output-feedback MPC into the tracking layer. This extension, called robust output-feedback hierarchical MPC (ROHMPC), provides formal safety guarantees even in the presence of disturbances and measurement noise. Synthesizing the ROHMPC scheme requires knowledge of uncertainty bounds, which are typically unknown.

To overcome this, the chapter introduces an efficient and modular pipeline that estimates these bounds from experimental data, performs necessary offline computations, calibrates constraint tightening to reduce conservatism, and implements the complete control scheme. The pipeline is released as an open-source software package to support reproducibility and future research. Using this pipeline, the chapter demonstrates the successful validation of the ROHMPC framework properties on a simulated quadcopter platform in Gazebo, with reproducible results.

The HMPC and ROHMPC frameworks address autonomy and safety respectively, but successful deployment in real-world scenarios also depends on the reliability of the system. A key aspect of reliability is reproducibility: the ability to consistently generate similar results.

Chapter 5 explores this concept in the context of robotics, defining reproducibility and analyzing how it applies to the hardware-software setups used in earlier chapters. The HMPC framework satisfies method reproducibility, meaning its implementation can be consistently reproduced. However, due to asynchronous processes and non-deterministic code components, it does not fully achieve results reproducibility.

In contrast, the ROHMPC framework satisfies both method and results reproducibility, reinforcing the credibility of the framework. By raising awareness of reproducibility challenges and offering practical insights, this chapter aims to support the development of more robust and trustworthy robotic systems.

All results presented in this thesis have been made publicly accessible through submissions to peer-reviewed venues, an open-access preprint server, and the release of open-source software packages. These results highlight the effectiveness of hierarchical MPC in both simulation and laboratory settings, and demonstrate how formal safety guarantees can be achieved under uncertainty.

To support future research, Chapter 6 summarizes the key contributions and outlines several promising directions for further exploration. These include extending the proposed algorithms to 3D environments, integrating onboard sensing for autonomous outdoor navigation, and incorporating data-driven methods to reduce conservatism. ... Read more

This thesis presents a comparative analysis of multi-target tracking algorithms for estimating the kinematic state and physical extent of maritime vessels from 4D millimetre-wave radar point clouds. The research addresses a gap in understanding the trade-offs between algorithmic complexity and motion model fidelity for inland maritime navigation.
The study evaluates three tracking paradigms, Tracking-by-Detection (TBD), the Probability Hypothesis Density (PHD) filter, and the Poisson Multi-Bernoulli Mixture (PMBM) filter, each tested with linear (Constant Velocity, Constant Acceleration) and non-linear (Constant Turn Rate) motion models. To represent vessel shape, three extended object models were implemented: the Random Matrix, Star-Convex Gaussian Process, and Principal Axis models.
Performance was evaluated using a mix of simulated datasets with perfect ground truth and collected real-world datasets. A hyperparameter optimisation process was conducted for every algorithm combination using the Optuna framework, with the objective of minimising the Generalised Optimal Sub-Pattern Assignment metric using a specified cost function.
The results yielded several key, and at times counterintuitive, findings. Firstly, increased algorithmic complexity did not yield superior performance. The simpler TBD and PHD frameworks consistently outperformed the theoretically optimal but computationally intensive PMBM filter, which was prone to generating unstable tracks and excessive false positives. Secondly, non-linear motion models offered no significant advantage over the simpler linear models. This could be attributed to the slow dynamics of maritime vessels, the high sensor update rate, which makes linear extrapolation sufficient, and the fact that measurement noise often dominates any subtle gains from a more precise motion model. Thirdly, for extent estimation, the Random Matrix model demonstrated the best balance of accuracy, stability, and computational efficiency. The more complex Gaussian Process and Principal Axis models struggled with stability and practicality.
The study concludes that the optimal balance is achieved by pairing simple linear motion models with the Random Matrix extent model, yielding accurate cardinality estimation, reliable state tracking, and efficient computation. The findings underscore that robustness and interpretability outweigh algorithmic sophistication when designing maritime tracking systems based on 4D radar. Ultimately, the thesis establishes a principled performance baseline and highlights the diminishing returns of complexity in this domain. ... Read more

Through the use of Eigenmanifold theory, unforced periodic trajectories called nonlinear normal modes can be identified and excited in nonlinear mechanical systems. Applying this to repetitive tasks in robotic systems with elastic components can drastically reduce energy consumption, as normal modes in steady-state require no additional control effort. However, existing control methods for exciting nonlinear normal modes have so far only assumed full actuation. Consequently, these techniques are incompatible with series elastic joint robots, even though they represent a significant subclass of physical systems with elastic elements. Additionally, the calculation and parameterization of Eigenmanifolds for high-dimensional systems generally remains a complex task and is difficult to scale. While existing literature aims to avoid forced evolutions or model cancellation, we instead lean into this approach. By rephrasing Eigenmanifold-based control as a trajectory tracking problem, standard techniques for elastic joint robot trajectory tracking control can be employed. Furthermore, obtaining theoretical guarantees on global stability becomes possible. In this work, a new modular control architecture is presented that integrates trajectory tracking feedback control with Eigenmanifold theory to dynamically generate and track hyper energy-efficient oscillatory movements in underactuated systems. This approach enables the excitation of nonlinear normal modes using standard trajectory tracking controllers, while preserving energy-efficient properties desired from Eigenmanifold-based controllers. We first discuss the theoretical validity and energy-efficiency of this control architecture, and then test the architecture in simulation for a variety of use-cases and controllers. ... Read more

Public safety and emergency response agencies increasingly consider the deployment of mobile robots as mounting climate-related disasters and security challenges place human personnel at higher risk and stress. Mobile robots, such as drones, are a promising strategy to respond to these challenges: They can navigate difficult, hazardous terrain, gather real-time situational data, and conduct search or reconnaissance tasks without putting humans at direct risk. However, the currently practiced teleoperation of robots is challenging for such complex missions since the simultaneous navigation, situation assessment, and search tasks can overload human cognitive abilities. Therefore, autonomous planning and decision-making algorithms are required to enable robots to explore and search unknown environments for targets such as missing persons or hazardous materials.

Moving towards this goal, this thesis addresses two core problems. First, local motion planning must carefully account for information gained from sensor observations as well as collision avoidance and the robot’s dynamics while moving through cluttered, unknown areas. Second, global exploration planning must strategically select where in the environment to explore to find the target quickly - especially when the environment is large or complex. Given that human operators often possess semantic knowledge about likely target locations, we hypothesize that incorporating such guidance by observed semantic features (e.g., object or room types) into the exploration planning is crucial for time-efficient autonomous search. We address these two core problems by making the following contributions.

The first contribution of the thesis is an informative local motion planning approach
that generates safe, collision-free trajectories around obstacles while minimizing uncertainty about the target locations. The critical challenge is to achieve computationally efficient planning of trajectories that maximize information gain under the robot’s kinodynamic constraints. In the proposed approach, a model predictive control (MPC) motion planner is guided by a learned viewpoint policy. The policy is trained via deep reinforcement learning (DRL) to maximize long-term information gain by providing a local subgoal to the MPC. The MPC follows the subgoal and ensures that the motion plan remains feasible and collision-free. Therefore, the robot can rapidly replan safe and informative local trajectories online. Simulation experiments demonstrate that the method achieves competitive performance in locating targets compared to a computationally expensive state-of-the-art planner using Monte Carlo Tree Search (MCTS), while allowing for significantly faster execution and replanning.

While local informative planning is crucial for exploring cluttered spaces, it often be-
haves myopically and inefficiently with respect to large and complex environments. Therefore, the second contribution introduces a global target search planner that balances directed search towards semantically promising areas with complete search space coverage. This planner extends the idea of frontier exploration - focusing observations on the boundaries between explored and unexplored regions - to target search, where different frontiers are assigned a semantic priority. This priority represents the semantic relationships between the target and nearby objects. To minimize target search time, the target search planner schedules high-priority frontiers earlier by solving a custom combinatorial optimization problem to determine the visitation order. By integrating coverage gains into the frontier priorities, the planner ensures that the robot explores the environment efficiently while focusing on semantically relevant areas. We demonstrate this approach in two studies outlined below.

Large, high-quality datasets for learning target-specific semantic relationships are
scarce in many real-world scenarios, especially in search and rescue. The third contribution addresses this limitation by proposing a method to learn semantic priority models from expert feedback. Rather than collecting massive amounts of labeled data, the approach exploits an expert operator’s sparse guidance inputs in a few target search scenarios. This expert guidance selects a frontier to explore next, which is stored in a training dataset together with the frontier’s semantic features. An expert model is then trained to approximate a priority function that predicts how relevant each frontier is for the expert. By incorporating this learned priority function into the global target search planner, the robot can autonomously prioritize semantically relevant areas according to the expert’s semantic knowledge. Experiments show that using a small number of expert demonstrations is sufficient for the robot to significantly improve its search efficiency and reduce travel distance until the target is found.

Lastly, the thesis extends semantic target search to three-dimensional environments
by integrating it into a 3D planning pipeline for micro aerial vehicles (MAV). The pipeline first detects objects in the environment using onboard vision and associates them with priority values computed from pre-trained large language model (LLM) embeddings. These priorities are then propagated into frontiers in a 3D voxel map, indicating frontier regions that are most likely to contain the target. This enables the evaluation of frontier viewpoints for their information gain that accounts for both semantic priority and volumetric coverage. The viewpoint gains are then used in the combinatorial target search planner to prioritize the viewpoints that most likely lead to the target while ensuring efficient coverage of the environment. By integrating the MAV’s kinodynamic constraints into the planning costs, the system ensures smooth, feasible trajectories in real-time. Simulation studies reveal that semantically guided exploration leads to faster and more reliable target discovery than different purely coverage-based exploration baselines. Experiments with a real MAV in the lab confirm the approach’s ability to autonomously navigate an MAV through complex 3D environments to a target, exploiting semantic cues, maximizing
coverage, and avoiding collisions.

In summary, this thesis demonstrates how planning and learning techniques can be
combined for autonomous target search and exploration. These techniques enable mobile robots to navigate unknown environments efficiently and safely while searching for targets and collecting required information. Crucially, our proposed method for semantically guided frontier planning bridges the gap between recent learning-based navigation approaches and established planning-based approaches suitable for real-world robotic systems. By integrating semantic knowledge into robotic exploration, the proposed methods can reduce human operator cognitive load and, therefore, facilitate robot deployment in scenarios such as search and rescue or reconnaissance missions. ... Read more

As we increasingly strive to integrate robots into human-centric environments, safety is a top priority. Traditionally, rigid collaborative robots have relied on safety-aware computational control policies, which are susceptible to perception errors and often lead to overly cautious behavior that limits performance. In contrast, soft robotics offers a promising alternative by ensuring passive compliance throughout the robot’s structure via material softness. This mechanical compliance inherently mitigates safety issues arising from perception or control errors, although this has been paid with a substantial drop in precision. In recent years, significant advances have been seen in soft robotics, with exciting new developments in design, smart materials, actuators, sensors, models, and control strategies. However, the modeling and control of continuum soft robots continue to pose major challenges due to their infinite degrees of freedom, complex nonlinear dynamics, and time-dependent behaviors like hysteresis. As a result, soft robots often lack the necessary capability and motion precision, leading to a tradeoff where performance is sacrificed for safety. With this thesis, we argue that this tradeoff can be overcome by developing more advanced algorithms that can reason on the physics of the soft robot. More specifically, we propose combining powerful learned models with efficient and effective model-based control approaches that allow for interpretability into the actions and admit stability guarantees.

Currently, two main approaches exist for controlling soft robots. The first employs model-based control using approximated physics-based models derived from first principles. The second directly learns control policies, primarily through reinforcement learning. Both strategies face notable limitations. Existing model-based controllers are unable to fully manage and eventually exploit the dynamics of soft robots because their underlying models inadequately capture complex behaviors, particularly how actuation and external interactions affect the robot’s deformation. Moreover, deriving these models requires extensive expert knowledge. Additionally, the combined complexity and uncertainty of the dynamics between a soft robot and its environment make it currently infeasible to develop comprehensive world models from first principles alone, thereby motivating the integration of machine learning approaches that can effectively leverage data-driven insights. Conversely, directly learning the controller — such as via reinforcement learning — lacks interpretability and stability guarantees while being highly sample inefficient, a significant drawback given the time-dependent material properties and limited lifespan of current soft robots.

In this thesis, we contend that combining learned models with model-based controllers presents a promising alternative that brings together the advantages of both approaches: expressive, data-driven models that require less expert knowledge paired with controllers that are both interpretable and provably stable. Although recent years have seen increased interest in leveraging learned models for control, most work in this area depends on computationally intensive optimal control methods, such as MPC, to optimize the actuation sequence with the learned model. However, the high computational cost of solving these optimal control problems limits the maximum control frequency during deployment, preventing us from fully exploiting the dynamic capabilities of soft robots. Instead, this thesis pursues closed-form controllers that utilize the physical structure of learned models within an energy-shaping framework. The main challenge here is to develop approaches that integrate such physical structures—specifically, kinetic and potential energy terms—into the learning of dynamical models for soft robots. Before addressing this main challenge, we first had to advance physics-based models derived from first principles and identify novel techniques to leverage them for control. On one hand, this clarified which physical priors were available for learning, while on the other hand, it inspired new ways to integrate model-based controllers with learned models. The thesis addresses this topic through several interconnected key contributions.

First, we argue that quantifying the safety of soft robots is crucial for designing and controlling them to ensure that the closed-loop system meets the specific safety requirements of their intended applications. To this end, we present the first safety metric for continuum soft robots, which assesses the safety of an integrated soft robot design by accounting for both its embodied and computational intelligence.

Secondly, this thesis enhances shape sensing for soft robots by leveraging insights from kinematic models. We accomplish this by formulating and solving nonlinear optimization problems that align sensor measurements with the backbone shapes predicted by the kinematic model. We present two distinct approaches that integrate commercial sensors—namely visual and magnetic—with SLAM algorithms and a learned sensor measurement model, respectively, to accurately estimate the soft robot’s state, a key requirement for effective feedback control.

Thirdly, this thesis introduces advanced physics-based actuation models, including those for robots actuated by auxetic metamaterials - referred to as HSA robots—and models that capture the actuation dynamics of piston-driven pneumatic soft robots. We then leverage the acquired model insights to design provably stable nonlinear controllers—specifically, an integral-saturated PID combined with potential shaping and Cartesian-space impedance control for planar GSA robots, as well as a backstepping controller for pneumatic piston-driven soft robots. This contribution deepens our understanding of actuation, a critical aspect of soft robot behavior, and demonstrates how such insights can be incorporated into model-based control strategies. Moreover, experiments with HSA robots have highlighted the limitations of purely physics-based models in capturing complex phenomena like hysteresis, thereby motivating the exploration of learning-based approaches. In the future, the developed actuation models can serve as valuable physical priors for learned models.

Fourthly, the thesis presents techniques for learning soft robot models that incorporate physical structures while ensuring stability. We accomplish this by embedding physics-based dynamical models into the learning algorithm, which determines the free parameters of the dynamics and optionally optimizes a coordinate transformation—such as encoding into latent space. Two notable approaches are introduced: (1) an algorithm that extracts low-dimensional soft robot strain models from samples of the robot backbone’s shape evolution, and (2) a network of coupled harmonic oscillators for learning latent dynamics from high-dimensional observations like images. The explicit inclusion of kinematic and potential energy terms in these models allows for stability analysis using standard nonlinear system theory tools, such as Lyapunov methods. For instance, we prove that the coupled oscillator network is both globally asymptotically stable and input-to-state stable.

Fifthly, we exploit the physical structure of the learned models from contribution four to design closed-form setpoint regulators. The controller contains two key components: (1) a potential shaping feedforward term that positions the local/global minimum of the closed-loop potential energy at the setpoint by leveraging the learned model knowledge, and (2) an integral-saturated PID feedback term that rejects disturbances and compensates for modeling errors to prevent steady-state errors. The stability of the closed-loop system can then be analyzed using Lyapunov arguments.

Finally, the thesis explores methods for generating compliant motion behaviors in soft robots beyond low-level control. One approach focuses on assisting users, particularly elderly individuals, with activities of daily living by guiding the low-level controller with brain signals. This is achieved by combining motor imagery classification from wearable EEG devices with compliant impedance control in operational space. The second approach combines an orbitally stable dynamical system in latent space with a bijective neural network parametrized encoder to learn periodic motions from demonstrations. By avoiding reliance on time references, this learned motion policy enables natural and compliant tracking of demonstrated periodic motions. This contribution ensures that not just the robot structure and low-level controller are compliant, but also the high-level motion strategy. ... Read more

In urban environments like Amsterdam, where road congestion and stress on infrastructure are critical issues, the city’s waterways remain underutilized for transportation. Autonomous Surface Vessels (ASVs), making waterway transport cheaper and less labor intensive, present a potential solution by reducing transportation costs and alleviating road traffic. Although companies are developing ASVs, current solutions are mainly limited to larger, less congested waterways, with further innovation needed for denser urban areas.

The challenges in autonomous navigation for ASVs in urban canals stem from the complex nonlinear dynamics of vessels, which require long-term planning and rapid responses to environmental changes. Urban waterways are narrow and unstructured, with loosely defined navigation rules that can lead to discontinuities in the motion planner’s cost function. Typically, motion planners rely on predictions of other agents’ movements and plan around them, resulting in no interaction awareness. This approach can lead to the freezing robot problem in dense environments, where the ASV halts, deeming all the space unsafe. A local motion planner must address these challenges by ensuring long-horizon interaction-aware motions, rule adherence, and real-time planning.

An emerging approach in autonomous navigation is a sampling-based Model Predictive Control (MPC) strategy known as Model Predictive Path Integral control (MPPI). This algorithm approximates the optimal control sequence by sampling from a continuous input distribution. Thanks to its gradient-free nature, MPPI only requires collision checking to avoid collision and allows discontinuous cost functions. Additionally, its computation speed remains largely unaffected by the complexity of the robots’ nonlinear dynamics, enabling longer planning horizons. While this approach has grown in popularity in the motion planning community, its application in dynamic environments with multiple interacting agents remains relatively unexplored.

Building on the state-of-the-art in MPPI, this thesis first introduces an Interaction-Aware Model Predictive Path Integral (IA-MPPI) controller tailored for motion planning in crowded urban canals. While conventional planners passively react to other agents, IA-MPPI actively predicts and plans cooperatively in real-time, addressing the freezing robot problem and handling discontinuous navigation rules. The decentralized, communication-free architecture assumes agent cooperation and employs a two-stage sample evaluation to enhance efficiency. This approach manages nonlinear dynamics, exact obstacle shapes, and longer planning horizons, demonstrating robustness and efficiency compared to state-of-the-art MPC in multi-agent environments.

While IA-MPPI uses a constant velocity model for predicting other agents’ hidden goals, Chapter 4 of this thesis introduces a learning-based trajectory prediction model trained on realistic artificial data to improve accuracy in crowded environments. By extracting the agents’ goals from predicted trajectories and integrating this information into the IA-MPPI controller, the planner’s performance increases significantly in environments with tight interactions. Moreover, this approach outperforms treating the predicted trajectory as occupied space, leading to safer navigation in dense urban canals.

One of MPPI’s main drawbacks is that it struggles in highly dynamic environments because it usually only samples around a previous plan. When rapid changes occur, such as strong disturbances or an obstacle unexpectedly cutting off the path, the previous plan becomes invalid, and all the generated samples may lead to collisions. To address this, Chapter 5 introduces Biased-MPPI, incorporating multiple classic and learning-based ancillary controllers into the sampling distribution. While the mathematical derivations show that this introduces a bias, this significantly improves reactivity, safety, and robustness to local minima. Additionally, this approach reduces the required samples, making the controller more efficient regarding computational demand.

Lastly, Chapter 6 leverages the gradient-free nature of MPPI and applies it to a different domain involving contact-rich manipulation tasks, which are notoriously difficult for model-based planning. With a parallelizable GPU-based physics engine like IsaacGym, MPPI can efficiently roll out hundreds of input sequences in parallel, faster than in real-time, to approximate optimal control. This approach eliminates the need for explicit modeling of contact dynamics by exploiting the models embedded in the simulator, making it particularly suited for manipulation tasks like pushing and picking with both prehensile and non-prehensile manipulators. Experiments with real robots demonstrate the effectiveness of this method in handling discontinuous, contact-heavy environments.

Overall, this thesis explores key aspects of motion planning, particularly for ASVs, with extensions to ground robots and mobile manipulators. It advances the MPPI framework for autonomous navigation by adapting it for multi-agent systems and introducing biases through classic and learning-based ancillary controllers. Additionally, the thesis compares MPPI to MPC in navigation experiments and to optimization fabrics in manipulation tasks. The work presented promotes the use of MPPI, especially in domains where complex nonlinear dynamics and discontinuous costs complicate the use of real-time optimization-based MPC. ... Read more

Autonomous mobile robots, once confined to structured settings like warehouses, now operate in dynamic, human-centered environments such as hospitals and streets, where interactions with humans are unavoidable. This shift poses major challenges for local motion planning, requiring robots to navigate complex environments where human behavior is difficult to predict while prioritizing safety, adaptability, and social compliance. This thesis addresses these challenges by proposing solutions for interaction-aware, safe, socially compliant, and adaptive planning.

We first examine the standard approach of splitting local motion planning into two sub-problems: predicting human trajectories and solving constrained trajectory optimization to avoid collisions. While this approach provides safety guarantees and can handle uncertainties, it neglects the interactions between the robot and the humans. We address this by formulating a Model Predictive Control problem in which the robot’s action influences both its own state and the human states. Focusing on navigation among pedestrians, we leverage the interpretable and established Social Force Model (SFM) to model the human response dynamics to robot actions. By accounting for the robot’s influence on pedestrian behavior, we demonstrate that the robot can guide pedestrian behavior. However, a carefully designed cost function is essential to promote cooperation and prevent exploitation.

Next, we focus on the limitations of learning-based approaches that can address social compliance, specifically, Imitation Learning (IL). While IL enables the learning of socially compliant behaviors from demonstrations or observations, the resulting policy lacks formal safety guarantees. Safety filters based on, e.g., Control Barrier Functions (CBFs), adapt control inputs to ensure safety and can be combined with IL policies. While CBFs are an effective tool to certify safety, two challenges remain: constructing them for complex systems with input constraints and accounting for model uncertainties. We propose Robust Policy Control Barrier Functions, a method for constructing robust CBFs that guarantees safety under worst-case bounded disturbances. Furthermore, we present a practical approximation and demonstrate its effectiveness in simulation and hardware experiments.

Next, we focus on adaptability, specifically in data-driven pedestrian prediction models, which are crucial for the decoupled prediction and planning approach. While existing models are trained offline on general datasets, they may not reflect the behavior of pedestrians in the robot’s environment. To address this, we propose a self-supervised continual learning framework that refines models during deployment using online data from the robot’s perception pipeline, preserving prior knowledge through regularization and selective retraining. Experiments show improved performance compared to naive online training. Although we focus on pedestrian prediction, the approach could extend to other methods applying learning-from-observations.

In summary, this thesis integrates interaction awareness into decoupled motion planning by leveraging the established SFM, presents a method for constructing safety filters using practical approximations of robust CBFs, and develops a framework that addresses adaptability through self-supervised continual learning. Integrating the safety filter with a socially compliant, interaction-aware policy learned from observations and adapted online through continual learning offers a comprehensive solution.
This combination could pave the way for local motion planning that is interaction-aware, safe, socially compliant, and adaptive. ... Read more

Reinforcement learning has emerged as a promising approach for enabling robots to learn from interactions with their environments, without relying on predefined behaviors. However, robots face significant challenges when learning directly from real-world interactions. Real-world learning is time-consuming and resource-intensive, often requiring extensive data collection over long periods. Additionally, the risks involved in trial-and-error learning in physical settings are high, as faulty policies can lead to safety issues or system damage. Simulations offer a safer and more efficient alternative, allowing robots to learn in simulated environments at faster-than-real-time speeds. Despite these benefits, simulations often serve as imperfect approximations of reality. As a result, robots may learn behaviors that exploit simulation-specific quirks, which may not perform well in real-world settings, creating difficulties in transferring learned behaviors from simulation to real environments—a challenge known as the sim-to-real gap. Several factors contribute to the sim-to-real gap, such as unmodeled physical phenomena like friction and deformation, and the asynchronous nature of real-world systems that simulations often fail to capture accurately. Additionally, using separate software stacks for simulation and deployment can unintentionally lead to discrepancies. Finally, simulating at faster-than-real-time speeds with asynchronous frameworks that distribute computation across multiple cores may also introduce inaccuracies without proper synchronization.

This thesis focuses on improving simulation tools and methodologies to enhance the efficiency and effectiveness of learning-based approaches in robotics. The work addresses key trade-offs between flexibility, speed, and accuracy in robotic simulations, which are critical for successfully transferring learned policies from simulation to real-world environments. Additionally, it introduces a strategy to improve resilience, ensuring that learned behaviors are robust to irrelevant and unknown dynamics.
By tackling these challenges, this thesis provides insights into the design of effective robotic simulators and presents contributions that help bridge the gap between simulated and real-world robotic learning. ... Read more

Spatiotemporal systems are systems whose dynamics depend on time and space and are commonly found in real life. These systems are mathematically modeled using partial differential equations and are also known as distributed-parameter systems. Due to their structure and the high number of variables involved, control and estimation for this class of systems are very challenging. This thesis addresses two problems related to spatiotemporal systems: state estimation and system identification.

Monitoring the states of a control system is important to ensure the behavior of the system is achieving the control objectives. This can be achieved, among others, by using state observers that estimate the states of the systems regularly. First, we present a literature review of observer design methods for distributed parameter systems. In general, the design requires a dimension-reduction approach to implement the observer. From the dimension reduction, the design approaches can be classified into late and early lumping. In the late lumping perspective, model reduction is performed at the end of the observer design. In the early lumping perspective, dimension reduction is applied to the model of the system. We incorporate both approaches in our literature review.

State observer design requires the model of the systems. This thesis also presents a system identification method for distributed-parameter systems. The identification of such systems typically requires spatially dense and regular measurements, followed by selecting sensors that provide significant measurements to the model to reduce the model complexity. However, these requirements may be challenging to fulfill. In case the sensor locations are irregular and sparse in space, we propose the use of lumped-parameter system identification.

For models with a large number of regressors, we propose a method for reducing the number of regressors using a tree representation. The tree is a way to list models with different numbers of regressors. From all possible regressors for the model, the proposed method builds the tree from the simplest models, i.e., models with one regressor. The number of regressors in the models is incrementally increased to one or more models with the best performance. The addition is repeated until the tree contains models with the desired maximum number of regressors.

System identification is typically performed using a complete data set, i.e., for each input sample, there is an associated output sample available. However, there are cases in which some output samples are not recorded in the data set, making the identification data incomplete. This thesis also considers the problem of incomplete data for Takagi-Sugeno (TS) fuzzy system identification using the product space clustering method. This method comprises two steps: fuzzy clustering and rules construction. The first proposed method enables the use of incomplete system identification data to fuzzy c-means clustering algorithm developed for incomplete classification, which yields different estimates for a missing sample. This can be achieved by fusing those different values into a single value. The second proposed method treats missing samples as optimization variables during the identification process. The optimization is repeated until the change of all optimization variables is small.
... Read more

... Read more

Neural Cellular Automata (NCA) have recently been proposed as neuromorphic robot controllers. Despite their promising behavioural characteristics and small parameter counts, training NCA for control tasks has proven difficult and unstable. It is so tricky that curriculum-like multi-stage training programs must be used for simple control tasks. This thesis posits how criticality theory, an amalgam of statistical mechanics and neuroscience, presents a compelling case for pre-training NCA into a critical state. Mainly the increase in inter-cellular com-munication distance and maximization of available information. This thesis presents a novel NCA update function architecture loosely based on neuroscience and two novel interchangeable pre-training methods, one implicit and one explicit, based on criticality theory aimed at improving training performance. The new architecture and pre-training methods are tested on the Cart-Pole environment and trained with Double Deep Q-Learning (DDQN) and neuro-evolution. The new architecture improves the training speed and general performance of the NCA, whilst the two pre-training steps greatly stabilise the control task training when using DDQN. The explicit methods resulted in faster agent training than the implicit methods, but the pre-training step was prone to failure, whereas the implicit methods always succeeded. Neuro-evolution was more efficient than DDQN when training iterations and helped explain the dynamics that make NCA challenging to train. The architecture and pre-training steps were further tested on the LunarLander problem, a more complex control task. The neuro-evolution method succeeded in training but did not present excellent results, whilst DDQN failed outright to train. ... Read more

In this master thesis, the development of an innovative gripper specifically designed for grasping the delicate peduncles of vine tomato trusses is presented. This research addresses a critical challenge in agricultural automation: efficiently and safely manipulating high-value crops without causing damage. The thesis begins with a comprehensive review of the current state of agricultural automation, particularly focusing on the post-harvest handling of vine tomatoes.
Building on identified limitations in existing robotic grippers and automated systems, a novel gripper design is proposed. This design uniquely approaches grasping the peduncle of vine tomato trusses. Traditionally, trusses have been grasped by the peduncle with standard pinching grippers for pick and place operations. The proposed gripper in this paper grasps the truss with a hook around the peduncle, often optimally at the centre of mass, increasing the success rate of grasping, stability, and avoiding damage to the truss. Furthermore, it has a higher tolerance for detection errors, allowing for inaccurate positioning of the robotic system. A hook-gripper can successfully grasp a wider range of tomato varieties than a pinch gripper.
The hook, consisting of two fingers, allows for more stable lifting of the truss and can handle peduncles in hard-to-grasp positions, such as in a crate filled with tomatoes. In addition to increased reach capabilities, the hook-gripper also enables manipulations like dragging and pushing.
The study involved an iterative design process, prototyping, and rigorous testing of the gripper. Key features include a hook mechanism for secure grasping, enhanced mobility for reaching into cramped spaces like packed crates, and a delicate touch to prevent bruising or damaging the fruit. The research also integrates the gripper with advanced detection systems for precise and effective operation within automated setups.
Results from extensive testing demonstrate that the newly designed gripper not only improves the success rate of grasping and manipulating vine tomato trusses but also significantly reduces the risk of damage compared to conventional pinch grippers. Individual trusses can be grasped with great success.
Testing for the different positions showed an increased range of grasping position that resulted in a successful grip. In practical experiments, the gripper performed well, lifting trusses with ease. Other test results show that the position of the peduncle is of great importance for the success rate. Test results indicated an 80% success rate in grasping trusses positioned near the edges of the crate. Replaying waypoint with "learning from demonstration" improves the grasping of the trusses compared to existing detection possibilities, and that emptying a crate is a challenging task.
The specialized hook-gripper offers insights into a practical solution for picking and placing vine tomatoes using the peduncle. This thesis contributes to the field of agricultural robotics and facilitates a step towards future innovations in the automation of high-value crop handling. The study delves into hook-gripper design and actuation, crucial for optimal performance in robotic manipulation systems.

... Read more

Robots have the potential to assume tasks across various real-world scenarios. To achieve this, we require adaptable and reactive robots that can robustly deal with products and environments that present variability. For example, in the agro-food sector, each tomato plant inside a greenhouse is unique; hence, different robotic motions are required when interacting with different plants. Unfortunately, due to their simplicity, most robotic solutions currently employed are rigid and rely on hand-crafted rules. Such solutions perform well in controlled and repetitive environments; however, they fall short when these conditions are not met. As a consequence, a large family of problems remains unsolved..... ... Read more

Physical human-robot cooperation (pHRC) has the potential to combine human and robot strengths in a team that can achieve more than a human and a robot working on the task separately. However, how much of the potential can be realized depends on the quality of cooperation, in which awarenes of the partner’s intention and preferences plays an important role. Preferences tend to be highly personal, and additionally depend on the cooperation partner and the cooperation itself. They can be hard to define in terms a robot would understand, and may change over time. This thesis focuses on learning ‘useful models’ from observed behavior, to let our robot adapt its behavior to better match its human partner’s preferences, and thus improve the cooperation.
The aim is to capture personalized approximate models of human preferences –how a person likes to do something– from very few interactive observations, providing only small amounts of imprecise data, such that the robot can use the model to improve each user’s comfort. First, we learn a model to predict and optimize the human ergonomics in a pHRC task, such that our robot can ropose a plan, for both the human and itself, to solve the task in a way that is more ergonomic for its human partner. However, people do not necessarily prefer to act ergonomically, nor do we want to impose on them what a robot thinks best. Therefore, next, we apply inverse reinforcement learning (IRL), to capture less restrictive preference models: 1) path and velocity preferences for motion planning, and 2) on a higher level of abstraction, which (grasp or motion) action to initiate for proactive physical support. For learning to take the correct action in cooperation, we developed the disagreement-aware variable impedance (DAVI) controller to smoothly transition between providing active guidance and allowing the human to demonstrate alternative behavior..... ... Read more

Currently, vine tomato weighing and packaging require a significant amount of manual work. The main barrier to automation lies in the difficulty of developing a reliable robotic grasping system for grasping already harvested vine tomatoes. We propose a method to grasp harvested vine tomatoes that lay in a horizontal position with considerable clutter, which is how they are commonly stored in a crate after harvest. The method consists of a computer vision pipeline to first identify individual trusses in the crate and then determine a suitable grasping location on the stem. The robot then executes a pinch grasp on the stem without the need for touch sensors or mechanistic models. Lab experiments were conducted using a robotic manipulator equipped with an eye-in-hand RGB-D camera, which resulted in a success rate of 89.3%. Furthermore, due to the continuous, online learning capabilities of the system, the robot was successfully able to pick and relocate all presented vine tomatoes when allowed to retry after a failed attempt. Lastly, the method is tried on objects other than vine tomatoes by training on synthetic data, which shows promising results. ... Read more

State-of-the-art object grasping with 7-DOF robotic manipulators requires joint configuration planning methods in order to provide position control of the end-effector. These motion planners are able to calculate a motion plan to execute a safe grasp, while taking environmental constraints into account. In human-robot interaction, a well known problem is that humans are uneasy with the arm motion the robot executes, because the motion plan lacks parametrization of variables which would account for the impression of legibility. In this study we develop a method which allows for teleoperated learning from a single demonstration that is perceived more legible by humans. The operator uses the Geomagic Touch haptic device to demonstrate a movement of the robot’s end-effector. Modeling a motion path from a single teleoperated demonstration is achieved using Dynamic Movement Primitives. The effectiveness of the teleoperated LfD module has been demonstrated both in simulation and on a TIAGo robot in a variety of poses. An experiment is conducted in which a state-of-the-art motion planner was compared to the proposed LfD method and the ability of human participants to predict the goal object of the robot. Using the teleoperated LfD method, the ability to predict the goal objects increases significantly and the human is more confident in making the prediction (P = 0.0102 and P < 0.001, respectively). This means that with the learning method a more legible grasp was generated than with the state-of-the-art motion planner. ... Read more