580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How can heuristic and metaheuristic algorithms improve the efficiency and quality of job scheduling in complex and dynamic job shop environments?
This research area focuses on developing approximate solution methods, including heuristics and metaheuristics such as genetic algorithms, ant colony optimization, iterative flattening search, and fuzzy logic based systems, to address the challenges of job shop scheduling problems (JSSP) and their variants. Given the NP-hard nature of these problems and the complexity introduced by dynamics and flexibility, exact methods are often computationally infeasible for large-scale or realistic instances. Heuristics and metaheuristics thus provide scalable approaches capable of handling multiple objectives, dynamic job arrivals, machine flexibility, and practical constraints like due dates and resource availability, offering near-optimal solutions within reasonable computation times.
Key finding: This paper proposes two adaptive search strategies integrated into genetic programming (GP) that utilize the frequency of features in high-performing individuals to guide the search effectively. Applied to the Dynamic... Read more
Key finding: This study integrates Ant Colony and Genetic Algorithms into a generative scheduling method complemented by dispatching rules to solve a real-world job shop scheduling problem in textile manufacturing. By incorporating... Read more
Key finding: This work introduces hybrid Genetic Algorithm (GA) approaches enhanced with priority rules—partial reordering, gap reduction, and restricted swapping—to solve job shop scheduling problems aiming to minimize makespan. The... Read more
Key finding: The paper introduces an Iterative Flattening Search (IFS) metaheuristic that iteratively relaxes and solves subsets of scheduling decisions to minimize makespan in the Flexible Job Shop Scheduling Problem (FJSSP). Employing... Read more
Key finding: The paper develops a fuzzy logic-based priority rule system that integrates multiple parameters with varying degrees of importance to generate dynamic job shop schedules. By contrast to traditional priority rules that address... Read more
2. What system design considerations and architectures support the effective implementation of automated job scheduling in manufacturing and IT systems?
This theme encompasses research addressing the architecture, data management, user interaction, and integration challenges in building practical automated scheduling systems suitable for real-world manufacturing and IT environments. It examines database and knowledge-base designs, scheduling engine architectures, and user interface considerations necessary to manage complex scheduling scenarios with dynamic data and evolving constraints. The focus lies on improving usability, scalability, system responsiveness, and the incorporation of domain-specific knowledge for scheduling in industrial settings.
Key finding: This work analyzes the architecture of scheduling systems, emphasizing three primary components: database/knowledge base modules, scheduling engine modules, and user interface modules. It highlights the shift from traditional... Read more
Key finding: This article presents a high-level simulator modeling system-wide interactions in computer networks, focusing on job routing and scheduling as adaptive, machine learning-driven components. It argues for the necessity of... Read more
Key finding: The paper discusses simulation techniques integrated into an electronic planning board for finite capacity scheduling at the factory floor level. It stresses the dependency of effective simulation on shop floor data... Read more
Key finding: This paper describes the design of an Automated Guided Vehicle (AGV) system, named 'Mover', integrating shortest job first scheduling with shortest path routing via the A* algorithm to optimize job processing and material... Read more
3. How can genetic programming-based hyper-heuristics be used to automatically evolve adaptive and multi-objective dispatching rules for job shop scheduling?
This research theme investigates the application of genetic programming (GP) within hyper-heuristic frameworks to evolve dispatching rules and scheduling policies that effectively adapt to dynamic job shop environments and simultaneously optimize multiple conflicting objectives such as makespan, tardiness, and due date adherence. The focus is on methodological innovations that enhance GP representations, evaluation schemes, and co-evolution strategies to produce reusable, competitive, and interpretable scheduling heuristics without extensive manual design, targeting scalable automated heuristic design in job shop scheduling.
Key finding: The paper develops a genetic programming based hyper-heuristic (GPHH) approach to evolve dispatching rules for job shop scheduling that adapt to dynamic shop conditions and multiple objectives. It introduces novel... Read more
Key finding: This study proposes a genetic algorithm-based hyper-heuristic that selects among 26 different dispatching rules as low-level heuristics to solve job shop scheduling problems. Using a benchmark set from OR-Library, the... Read more