Catherine Cleophas

When Are Deliveries Profitable

The paper aims to optimize the final part of a firm’s value chain with regard to attended last-mile deliveries. It is assumed that to be profitable, e-commerce businesses need to maximize the overall value of fulfilled orders (rather than their number), while also limiting costs of delivery. To do so, it is essential to decide which delivery requests to accept and which time windows to offer to which consumers. This is especially relevant for attended deliveries, as delivery fees usually cannot fully compensate costs of delivery given tight delivery time windows. The literature review shows that existing order acceptance techniques often ignore either the order value or the expected costs of delivery. The paper presents an iterative solution approach: after calculating an approximate transport capacity based on forecasted expected delivery requests and a cost-minimizing routing, actual delivery requests are accepted or rejected aiming to maximize the overall value of orders given the computed transport capacity. With the final set of accepted requests, the routing solution is updated to minimize costs of delivery. The presented solution approach combines well-known methods from revenue management and time-dependent vehicle routing. In a computational study for a German metropolitan area, the potential and the limits of value-based demand fulfillment as well as its sensitivity regarding forecast accuracy and demand composition are investigated.

Recent developments in demand forecasting for airline revenue management

Revenue management for airlines strives to maximise revenue as the sum of fares earned through customer bookings. Since the internet enables customers to make more informed choices and low-cost carriers introduced the concept of restriction-free fares, forecasting for airline revenue management has faced new challenges. In addition, advances in dynamic pricing allow for a more sophisticated approach to pricing, but also ask for more information on customer reactions. This paper summarises recent developments in demand forecasting with regard to the prediction of demand in the airline industry. A classification by demand arrival, level, detruncation and behaviour is applied.

Assessing Multi-Agent Simulations : Inspiration through Application

The application of multi-agent simulations in practical decision support and training gains relevance as technological advances improve computational performance, user interfaces and visualizations. This paper describes the life cycle of one such application, the airline revenue management simulation system REMATE. It highlights the way in which issues of verification, validation and acceptance were treated when implementing and applying REMATE. Feedback loops linking the system, practitioners and researchers are illustrated. Challenges with regard to the required balance of parsimony and realism required from the underlying model are summarized and critically assessed. The paper suggests that through the diverging or additional requirements of practical application, challenges and opportunities for further research in the field of multi-agent simulations arise.

Genetic algorithms for calibrating airline revenue management simulations

Revenue management (RM) theory and practice frequently rely on simulation modeling. Simulations are employed to evaluate new methods and algorithms, to support decisions under uncertainty and complexity, and to train RM analysts. To be useful in practice, simulations have to be validated. To enable this, they are calibrated: model parameters are adjusted to create empirically valid results. This paper presents two novel approaches, in which genetic algorithms (GA) contribute to calibrating RM simulations. The GA emulate analyst influences and iteratively adjust demand parameters. In the first case, GA directly model analysts, setting influences and learning from the resulting performance. In the second case, a GA adjusts demand input parameters, aiming for the best fit between emergent simulation results and empirical revenue management indicators. We present promising numerical results for both approaches. In discussing these results, we also take a broader view on calibrating agent-based simulations.

Benchmarking filter-based demand estimates for airline revenue management

In recent years, revenue management research developed increasingly complex demand forecasts to model customer choice. While the resulting systems should easily outperform their predecessors, it appears difficult to achieve substantial improvement in practice. At the same time, interest in robust revenue maximization is growing. From this arises the challenge of creating versatile and computationally efficient approaches to estimate demand and quantify demand uncertainty. Motivated by this challenge, this paper introduces and benchmarks two filter-based demand estimators: the unscented Kalman filter and the particle filter. It documents a computational study, which is set in the airline industry and compares the estimators’ efficiency to that of sequential estimation and maximum-likelihood estimation. We quantify estimator efficiency through the posterior Cramér–Rao bound and compare revenue performance to the revenue opportunity. Both indicate that unscented Kalman filter and maximum-likelihood estimation outperform the alternatives. In addition, the Kalman filter requires comparatively little computational effort to update and quantifies demand uncertainty.

Integrated Planning of Order Capture and Delivery for Attended Deliveries in Metropolitan Areas

The ongoing boom in e-commerce increases the importance of profitable and customer-oriented delivery services. Particularly in metropolitan areas, the high population density offers great potential for e-commerce, while uncertain demand and traffic conditions increase planning uncertainty. This contribution focuses on e-commerce delivery fulfillment (e-fulfillment) for attended last-mile delivery services in metropolitan areas. As the customer needs to be present for deliveries of groceries, for example, a service time window has to be agreed upon already when a customer’s order is accepted. We consider service time windows as a scarce resource and as the critical interface between order capture and order delivery. To optimally utilize this scarce resource, we propose combining concepts of revenue management and vehicle routing to extend tactical and operational planning for e-fulfillment. We define the research problem and provide a perspective on integrated planning for attended deliveries. Furthermore, we present the design of a virtual laboratory to support benchmarking in e-fulfillment research. To ensure realistic experimental settings, we plan to incorporate real-world data provided by a major e-grocery in Germany.

Exogenous Capacity Changes in Airline Revenue Management: Quantifying the Value of Information

In airline revenue management, capacity is usually assumed to be fixed. However, capacity changes are common in practice. This contribution quantifies the value of information when systematically considering possible capacity changes in revenue optimization. It solves a stochastic model that anticipates capacity changes, given different levels of information. A computational study compares solution approaches with respect to the resulting revenue, seat load factor, and denied boarding.

A Note on Effective Code-Share Control in Practice

Code-sharing enables airlines to jointly market their inventory and constitutes a central component in alliance strategy. This article discusses a widely applied code-share revenue management process and quantifies the spread of code-sharing based on empirical data provided by Lufthansa German Airlines. We identify decentralization, heterogeneity, information asymmetry, and selfishness as main challenges to the implementation of effective code-share control in the industry. We examine the implications of these challenges from a business as well as from an information systems perspective, and exemplify them in the context of revenue management. Finally, we outline steps to improve code-share control and identify opportunities for future research.

Wann sind Lieferaufträge profitabel

ZusammenfassungDer Beitrag zielt auf die Optimierung von Frei-Haus-Belieferungen ab und betrachtet damit das letzte Teilstück der Wertschöpfungskette. Es wird angenommen, dass Handel im E-Commerce letztendlich nur profitabel sein kann, wenn der Gesamtwert der erfüllten Aufträge (im Vergleich zur Maximierung der Anzahl der Aufträge) maximiert wird. Dabei müssen auch die Kosten der Auslieferung berücksichtigt werden. Entscheidend ist, welche Lieferanfragen akzeptiert und welche Lieferzeitfenster welchen Kunden angeboten werden sollen. Diese Frage ist besonders für Frei-Haus-Belieferungen relevant, da Liefergebühren bei engen Lieferzeitfenstern üblicherweise die Lieferkosten nicht ganz kompensieren können. Der Literaturüberblick zeigt, dass die existierenden Auftragsannahmestrategien den Auftragswert oder die erwarteten Lieferkosten ignorieren. Der Beitrag präsentiert einen iterativen Lösungsansatz: Erst wird die erforderliche Transportkapazität unter Berücksichtigung einer Vorhersage von erwarteten Lieferanfragen abgeschätzt. Die erforderliche Transportkapazität fließt in ein kostenminimales Routing ein. Dann werden tatsächlich auftretende Lieferanfragen akzeptiert bzw. abgelehnt. Dabei wird der Gesamtwert der akzeptierten Lieferaufträge unter Berücksichtigung der verfügbaren Transportkapazität maximiert. Auf Basis der akzeptierten Lieferaufträge werden kostenminimale Ausliefertouren generiert. Der präsentierte Lösungsansatz kombiniert etablierte Verfahren der Erlössteuerung mit solchen der zeitabhängigen Tourenplanung. Es wird eine Simulationsstudie für einen deutschen Ballungsraum durchgeführt, um das Potenzial und die Grenzen der wertbasierten Auftragserfüllung sowie ihre Robustheit in Bezug auf Vorhersagegenauigkeit und Nachfragestruktur zu untersuchen.AbstractThe paper aims to optimize the final part of a firm’s value chain with regard to attended last-mile deliveries. It is assumed that to be profitable, e-commerce businesses need to maximize the overall value of fulfilled orders (rather than their number), while also limiting costs of delivery. To do so, it is essential to decide which delivery requests to accept and which time windows to offer to which consumers. This is especially relevant for attended deliveries, as delivery fees usually cannot fully compensate costs of delivery given tight delivery time windows. The literature review shows that existing order acceptance techniques often ignore either the order value or the expected costs of delivery. The paper presents an iterative solution approach: after calculating an approximate transport capacity based on forecasted are accepted or rejected expected delivery requests and a cost-minimizing routing, actual delivery requests aiming to maximize the overall value of orders given the computed transport capacity. With the final set of accepted requests, the routing solution is updated to minimize costs of delivery. The presented solution approach combines well-known methods from revenue management and time-dependent vehicle routing. In a computational study for a German metropolitan area, the potential and the limits of value-based demand fulfillment as well as its sensitivity regarding forecast accuracy and demand composition are investigated.

Wann sind Lieferaufträge profitabel? - Berücksichtigung des Auftragswertes und der Transportkapazität in der Tourenplanung für die "letzte Meile" in Ballungsräumen.

ZusammenfassungDer Beitrag zielt auf die Optimierung von Frei-Haus-Belieferungen ab und betrachtet damit das letzte Teilstück der Wertschöpfungskette. Es wird angenommen, dass Handel im E-Commerce letztendlich nur profitabel sein kann, wenn der Gesamtwert der erfüllten Aufträge (im Vergleich zur Maximierung der Anzahl der Aufträge) maximiert wird. Dabei müssen auch die Kosten der Auslieferung berücksichtigt werden. Entscheidend ist, welche Lieferanfragen akzeptiert und welche Lieferzeitfenster welchen Kunden angeboten werden sollen. Diese Frage ist besonders für Frei-Haus-Belieferungen relevant, da Liefergebühren bei engen Lieferzeitfenstern üblicherweise die Lieferkosten nicht ganz kompensieren können. Der Literaturüberblick zeigt, dass die existierenden Auftragsannahmestrategien den Auftragswert oder die erwarteten Lieferkosten ignorieren. Der Beitrag präsentiert einen iterativen Lösungsansatz: Erst wird die erforderliche Transportkapazität unter Berücksichtigung einer Vorhersage von erwarteten Lieferanfragen abgeschätzt. Die erforderliche Transportkapazität fließt in ein kostenminimales Routing ein. Dann werden tatsächlich auftretende Lieferanfragen akzeptiert bzw. abgelehnt. Dabei wird der Gesamtwert der akzeptierten Lieferaufträge unter Berücksichtigung der verfügbaren Transportkapazität maximiert. Auf Basis der akzeptierten Lieferaufträge werden kostenminimale Ausliefertouren generiert. Der präsentierte Lösungsansatz kombiniert etablierte Verfahren der Erlössteuerung mit solchen der zeitabhängigen Tourenplanung. Es wird eine Simulationsstudie für einen deutschen Ballungsraum durchgeführt, um das Potenzial und die Grenzen der wertbasierten Auftragserfüllung sowie ihre Robustheit in Bezug auf Vorhersagegenauigkeit und Nachfragestruktur zu untersuchen.AbstractThe paper aims to optimize the final part of a firm’s value chain with regard to attended last-mile deliveries. It is assumed that to be profitable, e-commerce businesses need to maximize the overall value of fulfilled orders (rather than their number), while also limiting costs of delivery. To do so, it is essential to decide which delivery requests to accept and which time windows to offer to which consumers. This is especially relevant for attended deliveries, as delivery fees usually cannot fully compensate costs of delivery given tight delivery time windows. The literature review shows that existing order acceptance techniques often ignore either the order value or the expected costs of delivery. The paper presents an iterative solution approach: after calculating an approximate transport capacity based on forecasted are accepted or rejected expected delivery requests and a cost-minimizing routing, actual delivery requests aiming to maximize the overall value of orders given the computed transport capacity. With the final set of accepted requests, the routing solution is updated to minimize costs of delivery. The presented solution approach combines well-known methods from revenue management and time-dependent vehicle routing. In a computational study for a German metropolitan area, the potential and the limits of value-based demand fulfillment as well as its sensitivity regarding forecast accuracy and demand composition are investigated.