arXiv subjects

Paolo Luigi Rinaldi

Publications and source records attributed to Paolo Luigi Rinaldi.

3 recordsLinked to original sources

arXiv2025

Improving Order with Queues

Given a sequence of $n$ numbers and $k$ parallel First-in-First-Out (FIFO) queues, how close can one bring the sequence to sorted order? It is known that $k$ queues suffice to sort the sequence if the Longest Decreasing Subsequence (LDS) of the input sequence is at most $k$. But, what if the number of queues is too small for sorting completely? - We give a simple algorithm, based on Patience Sort, that reduces the LDS by $k - 1$. We also show, that the algorithm is optimal, i.e., for any $L > 0$ there exists a sequence of LDS $L$ such that the LDS cannot be reduced below $L - k + 1$ with $k$ queues. - Merging two sorted queues is at the core of Merge Sort. In contrast, two sequences of LDS two cannot always be merged into a sequence of LDS two. We characterize when it is possible and give an algorithm to decide whether it is possible. Merging into a sequence of LDS three is always possible. - A down-step in a sequence is an item immediately followed by a smaller item. We give an optimal algorithm for reducing the number of down-steps. The algorithm is online. Our research was inspired by an application in car manufacturing.

cs.DS↗

arXiv2025

Optimizing Car Resequencing on Mixed-Model Assembly Lines: Algorithm Development and Deployment

The mixed-model assembly line (MMAL) is a production system used in the automobile industry to manufacture different car models on the same conveyor, offering a high degree of product customization and flexibility. However, the MMAL also poses challenges, such as finding optimal sequences of models satisfying multiple constraints and objectives related to production performance, quality, and delivery -- including minimizing the number of color changeovers in the Paint Shop, balancing the workload and setup times on the assembly line, and meeting customer demand and delivery deadlines. We propose a multi-objective algorithm to solve the MMAL resequencing problem under consideration of all these aspects simultaneously. We also present empirical results obtained from recorded event data of the production process over $4$ weeks following the deployment of our algorithm in the Saarlouis plant of Ford-Werke GmbH. We achieved an improvement of the average batch size of about $30\%$ over the old control software translating to a $23\%$ reduction of color changeovers. Moreover, we reduced the spread of cars planned for a specific date by $10\%$, reducing the risk of delays in delivery. We discuss effectiveness and robustness of our algorithm in improving production performance and quality as well as trade-offs and limitations.

eess.SY↗

arXiv2025

Algorithm Engineering of SSSP With Negative Edge Weights

Computing shortest paths is one of the most fundamental algorithmic graph problems. It is known since decades that this problem can be solved in near-linear time if all weights are nonnegative. A recent break-through by [Bernstein, Nanongkai, Wulff-Nilsen '22] presented a randomized near-linear time algorithm for this problem. A subsequent improvement in [Bringmann, Cassis, Fischer '23] significantly reduced the number of logarithmic factors and thereby also simplified the algorithm. It is surprising and exciting that both of these algorithms are combinatorial and do not contain any fundamental obstacles for being practical. We launch the, to the best of our knowledge, first extensive investigation towards a practical implementation of [Bringmann, Cassis, Fischer '23]. To this end, we give an accessible overview of the algorithm, discussing what adaptions are necessary to obtain a fast algorithm in practice. We manifest these adaptions in an efficient implementation. We test our implementation on a benchmark data set that is adapted to be more difficult for our implementation in order to allow for a fair comparison. As in [Bringmann, Cassis, Fischer '23] as well as in our implementation there are multiple parameters to tune, we empirically evaluate their effect and thereby determine the best choices. Our implementation is then extensively compared to one of the state-of-the-art algorithms for this problem [Goldberg, Radzik '93]. On the hardest instance type, we are faster by up to almost two orders of magnitude.

cs.DS↗