Another round with the newest scientific papers using GeneticSharp.

1) Optimized Work Schedule Generation in Construction Projects

Abstract: This work introduces the Project Scheduling Problem with Marked Activities (PSP/mark), an extension of the classical Project Scheduling Problem where activities are annotated with environment-related properties, such as loudness or location, to ensure compliance with a work plan. Work plans specify when activities with certain properties are allowed or forbidden. Additionally, this work defines the Work Schedule Generation Problem (WSGP) which seeks an optimal recombination of given work plans to minimize a complex cost function. To calculate the costs, a solution to the PSP/mark is utilized. An ASAP (As Soon As Possible) scheduling scheme is proposed for the PSP/mark, while two genetic encodings for WSGP solutions are introduced for use with an evolutionary algorithm. By adapting a dataset of real-world projects, four benchmarking datasets for the WSGP were created and subsequently used for evaluation of the genetic encodings. The results show improvements of 10% and 14% in comparison to a trivial baseline. The applicability of these results to real-word instances of the WSGP is pending. Therefore, the more significant contribution of this work lies in the general insights gained into the novel scheduling problems, opening up avenues of future work. (paper)

2) Multiple Autonomous Drone Mission Planning

Abstract: The goal of this thesis is to develop an application capable of planning missions for the increasing number of drones, minimizing the need for human interaction in mission planning. Unmanned X Vehicle meaning UAV, UUV, USV or UGV (UXV)s are becoming increasingly prevalent in both military and civilian applications worldwide, offering significant advantages over manned vehicles, such as reduced maintenance requirements, enhanced safety, and freedom from human limitations. The widespread adoption and ongoing investments in UXV technologies have made them more cost-effective, allowing users to deploy not just a few but dozens of drones. (paper)

3) Selecting Deconstruction Processes using Virtual Reality Comparisons

Abstract: Grand challenges such as achieving sustainability goals and managing aging infrastructure are creating an unprecedented demand for deconstruction. However, deconstructing aging infrastructure is inherently risky, repetitive, and costly, necessitating effective project planning. Virtual reality (VR) technology offers the potential for planners to improve deconstruction project risk, time, safety, and cost. We propose that planners leverage VR early in the planning phase to compare alternative feasible tools and processes. Planners can create low-fidelity models of various deconstruction process alternatives and collect metrics on their suitability with VR-capable three-dimensional (3D) game engines. We present and formalize a methodology for conducting comparisons of candidate deconstruction processes by modeling candidates in VR, conducting trials, and collecting analytics data from the VR engine. We present a case study that demonstrates our approach to a cutting and waste packing process for nuclear power plant (NPP) decommissioning. As a novel contribution in this paper, we show that VR simulations can efficiently produce detailed insights useful for critically analyzing and comparing deconstruction process alternatives. (paper)

4) Movement Simulation of a Handheld Device

Abstract: Web browsers give websites access to motion sensors on mobile devices such as phones and tablets. The shared sensor data can be exploited to track and identify users. The JShelter browser extension provides protection against such exploitation of motion data by passing fake values. However, these values simulate a stationary device, which can lead to detection of the simulation. The goal of this thesis is to create a simulation that will generate believable device motion in the hands of a human. Prior to the design, sensor data analysis and exploration of motion simulation methods were conducted. Sets of parameters generated by a genetic algorithm are used for motion generation. The resulting solution was incorporated into the JShelter extension and experiments showed good results and performance of the solution. (paper)

5) Using Genetic Algorithms for the Generation of Increasingly Challenging Terrain for Players to Navigate

Abstract: This research focuses on the development of a system that employs genetic algorithms (GA) that generate increasingly challenging terrain for a player to navigate. Our system is developed within a video game, where a player is required to get from a start point to an endpoint, with gameplay mechanisms based on the terrain acting as obstacles for the player. The results from a game experience survey for the ingame section show that players rated gameplay more negatively when the terrain was randomly generated than when it was generated by the GA. While players did not feel more challenged using the GA-generated terrain, they gave an overall more positive gameplay experience. A quantitative data analysis of data collected during gameplay indicates an increase as the rounds progressed, in the number of deaths, the time taken, and levels skipped when using the GA. In contrast, the randomly generated terrain shows no significant difference. (paper)

6) Approaches to Multi-Constraint Job Order Balancing

Abstract: In scheduling, not all processes can be scheduled equally and may present their own unique set of constraints. Solution approaches include meta-heuristics and exact methods. Two different approaches were chosen to generate schedules with constraints and compare their performance when implemented for a scheduling activity; Constraint Programming and the Genetic Algorithm. Quasi-experiments were conducted to evaluate the execution time and accuracy score of each solution using a dataset of 50 jobs. The baseline includes a completed scheduling of the jobs. The results indicate that the Genetic Algorithm solution offers the best results in terms of execution time and accuracy, exhibiting results comparable to the baseline. The Constraint Programming solution failed to find any optimal results, demonstrating lower accuracy compared to the Genetic Algorithm and the baseline. With the foundation laid by this study, further work may improve each model to a more usable degree. (paper)

7) Discrete and mixed-variable experimental design with surrogate-based approach

Abstract: xperimental design plays an important role in efficiently acquiring informative data for system characterization and deriving robust conclusions under resource limitations. Recent advancements in high-throughput experimentation coupled with machine learning have notably improved experimental procedures. While Bayesian optimization (BO) has undeniably revolutionized the landscape of optimization in experimental design, especially in the chemical domain, it is important to recognize the role of other surrogate-based approaches in conventional chemistry optimization problems. This is particularly relevant for chemical problems involving mixed-variable design space with mixed-variable physical constraints, where conventional BO approaches struggle to obtain feasible samples during the acquisition step while maintaining exploration capability. In this paper, we demonstrate that integrating mixed-integer optimization strategies is one way to address these challenges effectively. Specifically, we propose the utilization of mixed-integer surrogates and acquisition functions–methods that offer inherent compatibility with problems with discrete and mixed-variable design space. This work focuses on piecewise affine surrogate-based optimization (PWAS), a surrogate model capable of handling medium-sized mixed-variable problems (up to around 100 variables after encoding) subject to known linear constraints. We demonstrate the effectiveness of this approach in optimizing experimental planning through three case studies. By benchmarking PWAS against state-of-the-art optimization algorithms, including genetic algorithms and BO variants, we offer insights into the practical applicability of mixed-integer surrogates, with emphasis on problems subject to known discrete/mixed-variable linear constraints. (paper)