Summary

In the context of the “Process Mining in Logistics” research project between Vanderlande Industries, we are offering multiple Master projects on process mining on event data of large-scale material handling systems. The fundamental challenges addressed are size (logistics processes are a factor 10-100 larger than business processes), reliable performance analysis and process prediction.

We offer several topics that build on a rich stack of recently developed advanced techniques for data preparation, process mining, and process prediction.

Vanderlande

Vanderlande is the global market leader for value-added logistic process automation at airports and in the parcel market. The company is also a leading supplier of process automation solutions for warehouses. Some figures:

Vanderlande’s baggage handling systems move 3.7 billion pieces of luggage around the world per year.
Our systems are active in 600 airports including 13 of the world’s top 20.
More than 39 million parcels are sorted by its systems every day, which have been installed for the world’s leading parcel companies.
More than 206 projects in 105 countries
Many of the largest global e-commerce players and distribution firms have confidence in Vanderlande’s efficient and reliable solutions.

Vanderlande focuses on the optimization of its customers’ business processes and competitive positions. Through close cooperation, we strive for the improvement of our customers’ operational activities and the expansion of their logistical achievements.

For Vanderlande, it is critical that we have state-of-the-art techniques to analyze and optimize our customers’ logistics processes. Reasons are (a) the constantly increasing size and complexity of our material handling solutions, (b) growing complexity of our software solutions, covering larger and larger parts of our customers’ business processes, and (c) the demand for more advanced service offerings, covering logistics and business services together. We believe that process mining is of high value for Vanderlande. Therefore, we work with the Eindhoven University of Technology on making process mining fit for analyzing logistics processes. In this context, we offer graduation positions on process mining and the application in our business.

Offered Topics

Analyzing Causes of Outlier Cascade Behavior in Baggage Handling Systems

Baggage Handling Systems have the task of transporting and processing bags from check-in, via various stages of scanning and storage, to the aircraft. Thereby all bags have to reach their aircraft at the scheduled moment in time to be loaded in the aircraft. The bags are transported via conveyor belts and other mechanical equipment. The usage and performance of this equipment by the bags passing over it determines whether the bag will reach its destination on time. In case of equipment failures or malfunctions, bags at the equipment cannot be transported further and suffer delays in processing until the equipment is functional again (called a dynamic bottleneck). But also further bags on conveyor belts right before the malfunctioning equipment are stuck, similar to traffic jam in which more and more incoming bags may get stuck (called a die-back effect).

In recent groundbreaking research (https://doi.org/10.1109/ICPM49681.2020.00014), we developed a technique to automatically detect system-level cascades of outlier behavior that lead to dynamic bottlenecks and die-backs. The technique aggregates event data of all bags in various parts of the system and identifies patterns between various system parts and their possible cause-effect relations. The technique however does not take the status of the equipment itself into account, stored in so-called SCADA reports. It is currently unclear how the dynamic bottlenecks that are detected relate to hardware sensor information. Aligning both information sources would allow validating and improving the dynamic bottleneck detection and would give vital early warning information about the likelihood that a dynamic bottleneck or die back may occur.

The objective of this project is to enhance the existing outlier cascade detection technique to also consider equipment status and failure events from the SCADA reports as features and events in the cascades. This requires developing a more bottleneck detection model that considers multiple information sources, developing a more general cascade model, and developing and implementing a scalable technique to detect these enhanced outlier cascades. The technique then shall be applied, evaluated, and improved regarding its ability to

Validate and improve the dynamic bottleneck detection from various information sources.
Detect cascades of SCADA events and system-level outlier behavior which give insights into propagation of equipment failures to identify SCADA events that serve as early warning detectors for dynamic bottlenecks and die-backs
Detect critical SCADA events which can serve as reliable early warning detectors for dynamic bottlenecks and die-backs

The expected outcome is a method and proof-of-concept implementation for detecting the Top-10 outlier cascades in standardized data of any Vanderlande baggage handling system. The project is offered in the context of a major international airport with data available in standard formats and access to process engineers familiar with the system and the problem.

Understand Bottlenecks and Performance Problems in Managing Tubs in Baggage Handling Systems

Modern Baggage Handling systems transport bags partly in so-called tubs - medium-sized open containers that can take in one bag and that are transported via conveyor belts. However, bags are not placed in tubs throughout the entire system: there are various points where a bag has to be placed into an empty tub or taken out of a tub. The empty tubs are moved forward to a location where they are needed next. Efficient system performance requires that empty tubs are managed correctly: whenever a bag needs to be placed into a tub an empty tub has to be available at the location where the bag needs it. If the tub is not available, the bag has to wait until the tub arrives. A system-wide control software predicts where tubs are needed next and dynamically routes empty tubs to the next location where they are needed.

The objective of this project is to provide insights into how the tub management and routing can be improved to reduce waiting time of bags for empty tubs, and to reduce the number of unnecessary tub movement in the system. Data about about baggage movement and tub movement is available in standardized data format. The goal is to develop a technique to

Integrate both datasets to understand and reason about baggage movement and tub movement in an integrated way through the use of graph-based data models and graph databases,
To develop a technique to visualize tub and bag movement and to visualize and automatically detect possible problems in tub management

The expected outcome of this project is a pipeline for integrating, processing, and visualizing the baggage and tub movement data, and to evaluate the visualization in its feasibility to identify problems in tub management.