Novel techniques in evolutionary optimization, simulation, and machine learning enable a broad analysis of domains like fluid dynamics, in which computation is expensive and flow behavior is complex. This paper introduces the concept of full domain analysis, defined as the ability to efficiently determine the full space of solutions in a problem domain and analyze the behavior of those solutions in an accessible and interactive manner. The goal of full domain analysis is to deepen our understanding of domains by generating many examples of flow, their diversification, optimization, and analysis. We define a formal model for full domain analysis, its current state of the art, and the requirements of its sub-components. Finally, an example is given to show what can be learned by using full domain analysis. Full domain analysis, rooted in optimization and machine learning, can be a valuable tool in understanding complex systems in computational physics and beyond.

Unmanned Aerial Vehicles (UAVs) are increasingly used for reforestation and forest monitoring, including seed dispersal in hard-to-reach terrains. However, a detailed understanding of the forest floor remains a challenge due to high natural variability, quickly changing environmental parameters, and ambiguous annotations due to unclear definitions. To address this issue, we adapt the Segment Anything Model (SAM), a vision foundation model with strong generalization capabilities, to segment forest floor objects such as tree stumps, vegetation, and woody debris. To this end, we employ parameter-efficient fine-tuning (PEFT) to fine-tune a small subset of additional model parameters while keeping the original weights fixed. We adjust SAM's mask decoder to generate masks corresponding to our dataset categories, allowing for automatic segmentation without manual prompting. Our results show that the adapter-based PEFT method achieves the highest mean intersection over union (mIoU), while Low-rank Adaptation (LoRA), with fewer parameters, offers a lightweight alternative for resource-constrained UAV platforms.

Effectively identifying and locating plants on the forest floor is essential for successful reforestation efforts and forest health assessment. This task is challenging due to the diverse range of plants, the high variance in appearance, the limited availability of accurately labeled data, and the complexity of data collection in post-harvest forest areas. These challenges are addressed by integrating field-surveyed plant data with images captured by Unmanned Aerial Vehicles (UAVs) to produce labeled training data. A custom Faster-RCNN model is trained using the aforementioned data to detect individual plants and clustered plant groups. Two different annotation approaches (unified class and distinct classes) are explored to address plant groupings and compare their performance using Mean Average Precision at 50% overlap (mAP50) and F1 score. The model trained using the unified class approach shows promise in detecting plants and plant groups larger than 0.15 square meters (F1 score: 0.44) but struggles with those smaller than 0.15 square meters. The inclusion of field-labeled data makes detection more challenging but ensures reliability. Meanwhile, the distinct class approach shows limited success. This work underscores the value of high-resolution imagery and comprehensive temporal analysis in enhancing detection accuracy and reliability.

Over the past few years, forests have become increasingly susceptible to various natural stress factors, mainly due to climate change induced events such as drought. Quantifying the objects that lay on the forest floor such as soil, vegetation, wooden debris and tree stumps is a viable method for estimating the extent of the stress factors. Remote sensing is an established modus operandi for collecting data in forest environments. In this work, we focus on Unmanned Aerial Vehicles (UAV) as remote sensing platforms aiming to detect tree stumps in varying stages of decay within a forest calamity area. The platform provides us with both high-resolution images from RGB and multispectral cameras and height data. In addition, a vegetation index for wood is proposed to better highlight the presence of tree stumps using multispectral data. Object detection is performed with feature-level fusion after which Object-Based Image Analysis (OBIA) is carried out. The fused multispectral and RGB data resulted in 53.68% recall and 31.0% precision in detecting tree stumps. In comparison, the fused RGB and height data resulted in 50.0% recall and 20.9% precision.

Climate change induced events such as drought lead to stress in forest trees, causing major catastrophic outbreaks of bark beetle insects. Current efforts to mitigate the infestation is to mass log an affected area. As a result the forest floor of a post-harvest area is cluttered with various objects such as tree logs and twigs. This study is aimed towards exploring basic computer vision methods that make use of shape, elevation and color to detect and segment objects on the forest floor. Such methods have advantages over learning methods due to their simplicity and speed in producing initial usable results, in addition to the low requirement of computational resources and training data. The highest intersection over union result of the multi-class detection and segmentation method is 0.64 for the road class. Such methods prove the feasibility of deploying basic computer vision techniques to acquire fast and reliable results.

Quality diversity algorithms can be used to efficiently create a diverse set of solutions to inform engineers' intuition. But quality diversity is not efficient in very expensive problems, needing hundreds of thousands of evaluations. Even with the assistance of surrogate models, quality diversity needs hundreds or even thousands of evaluations, which can make its use infeasible. In this study, we try to tackle this problem by using a pre-optimization strategy on a lower-dimensional optimization problem and then map the solutions to a higher-dimensional case. For a use case to design buildings that minimize wind nuisance, we show that we can predict flow features around 3D buildings from 2D flow features around building footprints. For a diverse set of building designs, by sampling the space of 2D footprints with a quality diversity algorithm, a predictive model can be trained that is more accurate than when trained on a set of footprints that were selected with a space-filling algorithm like the Sobol sequence. Simulating only 16 buildings in 3D, a set of 1,024 building designs with low predicted wind nuisance is created. We show that we can produce better machine learning models by producing training data with quality diversity instead of using common sampling techniques. The method can bootstrap generative design in a computationally expensive 3D domain and allow engineers to sweep the design space, understanding wind nuisance in early design phases.

Force field-based models are a Newtonian mechanics approximation of reality and are inherently noisy. Coupling models from different molecular scale domains (including single, gas-phase molecules up to multimolecule, condensed phase ensembles) is difficult, which is also the case for finding solutions that transfer well between the scales. In this contribution, we introduce a surrogate-assisted algorithm to optimize Lennard-Jones parameters for target data from different scale domains to overcome the difficulties named above. Specifically, our approach combines a surrogate-assisted global evolutionary optimization method with a presampling phase that takes advantage of one scale domain being less computationally expensive to evaluate. The algorithm's components were evaluated individually, elucidating their individual merits. Our findings show that the process of parametrizing force fields can significantly benefit from both the presampling method, which alleviates the need to have a good initial guess for the parameters, and the surrogate model, which improves efficiency.

This paper explores the role of artificial intelligence (AI) in elite sports. We approach the topic from two perspectives. Firstly, we provide a literature based overview of AI success stories in areas other than sports. We identified multiple approaches in the area of Machine Perception, Machine Learning and Modeling, Planning and Optimization as well as Interaction and Intervention, holding a potential for improving training and competition. Secondly, we discover the present status of AI use in elite sports. Therefore, in addition to another literature review, we interviewed leading sports scientist, which are closely connected to the main national service institute for elite sports in their countries. The analysis of this literature review and the interviews show that the most activity is carried out in the methodical categories of signal and image processing. However, projects in the field of modeling & planning have become increasingly popular within the last years. Based on these two perspectives, we extract deficits, issues and opportunities and summarize them in six key challenges faced by the sports analytics community. These challenges include data collection, controllability of an AI by the practitioners and explainability of AI results.

Agricultural fields suffer from damage due to changing climate conditions and wildlife foraging. Damage incurred by wild boar is classified as the main contributing factor to grasslands damage. Such damage results in losses for farmers due to reduced yield potential and repair costs. The extent of wild boar damage is typically estimated manually by using ground based approaches (e.g. GPS measurements), which is a time consuming task with questionable accuracy. Building upon our previous work, we present an autonomous approach to detect and classify the cause of damage to grasslands (wild boar, mole etc.). The approach entails utilizing convolutional neural networks for semantic segmentation of grasslands. An RGB baseline was established, in addition to evaluating multimodal architectures that incorporate different surface model feature representations, leading to a joint representation of spectral and elevation information. Testing and experimentation was performed in real-world grasslands around Bonn, Germany. The results show that incorporating elevation features with late fusion enhances the overall performance of the network over the RGB baselines.

We consider multi-solution optimization and generative models for the generation of diverse artifacts and the discovery of novel solutions. In cases where the domain's factors of variation are unknown or too complex to encode manually, generative models can provide a learned latent space to approximate these factors. When used as a search space, however, the range and diversity of possible outputs are limited to the expressivity and generative capabilities of the learned model. We compare the output diversity of a quality diversity evolutionary search performed in two different search spaces: 1) a predefined parameterized space and 2) the latent space of a variational autoencoder model. We find that the search on an explicit parametric encoding creates more diverse artifact sets than searching the latent space. A learned model is better at interpolating between known data points than at extrapolating or expanding towards unseen examples. We recommend using a generative model's latent space primarily to measure similarity between artifacts rather than for search and generation. Whenever a parametric encoding is obtainable, it should be preferred over a learned representation as it produces a higher diversity of solutions.

More and more, optimization methods are used to find diverse solution sets. We compare solution diversity in multi-objective optimization, multimodal optimization, and quality diversity in a simple domain. We show that multiobjective optimization does not always produce much diversity, multimodal optimization produces higher fitness solutions, and quality diversity is not sensitive to genetic neutrality and creates the most diverse set of solutions. An autoencoder is used to discover phenotypic features automatically, producing an even more diverse solution set with quality diversity. Finally, we make recommendations about when to use which approach.

Computers can help us to trigger our intuition about how to solve a problem. But how does a computer take into account what a user wants and update these triggers? User preferences are hard to model as they are by nature vague, depend on the user’s background and are not always deterministic, changing depending on the context and process under which they were established. We pose that the process of preference discovery should be the object of interest in computer aided design or ideation. The process should be transparent, informative, interactive and intuitive. We formulate HyperPref, a cyclic co-creative process between human and computer, which triggers the user’s intuition about what is possible and is updated according to what the user wants based on their decisions. We combine quality diversity algorithms, a divergent optimization method that can produce many, diverse solutions, with variational autoencoders to both model that diversity as well as the user’s preferences, discovering the preference hypervolume within large search spaces.

In complex, expensive optimization domains we often narrowly focus on finding high performing solutions, instead of expanding our understanding of the domain itself. But what if we could quickly understand the complex behaviors that can emerge in said domains instead? We introduce surrogate-assisted phenotypic niching, a quality diversity algorithm which allows to discover a large, diverse set of behaviors by using computationally expensive phenotypic features. In this work we discover the types of air flow in a 2D fluid dynamics optimization problem. A fast GPU-based fluid dynamics solver is used in conjunction with surrogate models to accurately predict fluid characteristics from the shapes that produce the air flow. We show that these features can be modeled in a data-driven way while sampling to improve performance, rather than explicitly sampling to improve feature models. Our method can reduce the need to run an infeasibly large set of simulations while still being able to design a large diversity of air flows and the shapes that cause them. Discovering diversity of behaviors helps engineers to better understand expensive domains and their solutions.

A probabilistic model of maximal mean performance in endurance sports is presented. The joint distribution of three variables namely interval length, average power, and average heart rate is modeled using Gaussian processes. The model allows for prediction of maximal average performances even based on data from sub-maximal efforts.

The encoding of solutions in black-box optimization is a delicate, handcrafted balance between expressiveness and domain knowledge -- between exploring a wide variety of solutions, and ensuring that those solutions are useful. Our main insight is that this process can be automated by generating a dataset of high-performing solutions with a quality diversity algorithm (here, MAP-Elites), then learning a representation with a generative model (here, a Variational Autoencoder) from that dataset. Our second insight is that this representation can be used to scale quality diversity optimization to higher dimensions -- but only if we carefully mix solutions generated with the learned representation and those generated with traditional variation operators. We demonstrate these capabilities by learning an low-dimensional encoding for the inverse kinematics of a thousand joint planar arm. The results show that learned representations make it possible to solve high-dimensional problems with orders of magnitude fewer evaluations than the standard MAP-Elites, and that, once solved, the produced encoding can be used for rapid optimization of novel, but similar, tasks. The presented techniques not only scale up quality diversity algorithms to high dimensions, but show that black-box optimization encodings can be automatically learned, rather than hand designed.

Mathematical models are simplifications of the real systems. The value of a mathematical model lies in its ability to reasonably model real-world phenomena. Cellular automata (CA) are a sub-class of agent-based models (ABM) that have been quite successful in modelling interacting real systems such as traffic flows, crowd behaviour, ecology, epidemics, chemical systems, etc. However, in the literature, CA models have been criticized as oversimplifications of reality and those that have relaxed rules have been criticized as not pure CA. In this paper, we show a procedure for calibrating a probabilistic-CA (P-CA) model to real system.

The initial phase in real world engineering optimization and design is a process of discovery in which not all requirements can be made in advance, or are hard to formalize. Quality diversity algorithms, which produce a variety of high performing solutions, provide a unique chance to support engineers and designers in the search for what is possible and high performing. In this work we begin to answer the question how a user can interact with quality diversity and turn it into an interactive innovation aid. By modeling a user's selection it can be determined whether the optimization is drifting away from the user's preferences. The optimization is then constrained by adding a penalty to the objective function. We present an interactive quality diversity algorithm that can take into account the user's selection. The approach is evaluated in a new multimodal optimization benchmark that allows various optimization tasks to be performed. The user selection drift of the approach is compared to a state of the art alternative on both a planning and a neuroevolution control task, thereby showing its limits and possibilities.

This work addresses the issue of finding an optimal flight zone for a side-by-side tracking and following Unmanned Aerial Vehicle(UAV) adhering to space-restricting factors brought upon by a dynamic Vector Field Extraction (VFE) algorithm. The VFE algorithm demands a relatively perpendicular field of view of the UAV to the tracked vehicle, thereby enforcing the space-restricting factors which are distance, angle and altitude. The objective of the UAV is to perform side-byside tracking and following of a lightweight ground vehicle while acquiring high quality video of tufts attached to the side of the tracked vehicle. The recorded video is supplied to the VFE algorithm that produces the positions and deformations of the tufts over time as they interact with the surrounding air, resulting in an airflow model of the tracked vehicle. The present limitations of wind tunnel tests and computational fluid dynamics simulation suggest the use of a UAV for real world evaluation of the aerodynamic properties of the vehicle's exterior. The novelty of the proposed approach is alluded to defining the specific flight zone restricting factors while adhering to the VFE algorithm, where as a result we were capable of formalizing a locally-static and a globally-dynamic geofence attached to the tracked vehicle and enclosing the UAV.

An analytical convolution-based model is used to predict a person’s physiological reaction to strain. Heart rate, oxygen uptake, and carbon dioxide output serve as physiological measures. Cycling ergometer tests of five male subjects are used to compare the proposed Convolution Model with a machine learning approach in form of a black box Wiener model. In these experiments, the Convolution Model yields smaller errors in prediction for all considered physiological measures. It performs very similar to other analytical models, but is based on only four parameters in its original form. A parameter reduction to one single degree of freedom is shown with comparable prediction accuracy and without significant loss of fitting accuracy.