Research projects

PhD's

Our research is to extend this work further to decentralised networks where required computations can be performed on the data stored on a blockchain.

Made by TU Delft's PhD students and directed by Zeki Erkin

A Privacy Preserving Multi-Depot Vehicle Routing Protocol
by M. Aysen

Decreasing costs may play a crucial role in small and medium-sized enterprises’ (SMEs) growth. To decrease transportation costs, these companies may form a collaboration for order sharing. Collaboration is crucial for the customer orders that they cannot fulfil in time or fulfil with low vehicle utilization. Various studies prove the significance of order sharing. However, the privacy of the data that are shared during this process does not take into account. In this work, we describe a privacy-preserving protocol for order sharing. We work on a case where each company has a single depot in a different location, formally named multi-depot vehicle routing problem (MD-VRP). The customer orders are distributed among companies by cyclic assignment, with the aim of minimizing the total traveled distance. In our protocol, we use threshold cryptography based on Paillier to make sure no single company can use the data based on their interest. By homomorphic properties and secure comparison protocol, we make sure the third party that executes the cyclic assignment does not learn anything about the data.

Solving bin-packing problems with privacy-preserving data publishing and optimization
by T. Li

An application domain where data play a crucial role is the container management system for transportation. In the largest ports around the world, thousands of container per day are being transported. Trucks bring containers in and out, and while doing so, it is commercially very important to use the container space as much as possible. To utilize the container space efficiently, optimization algorithms are proposed to arrange the packages in containers. While doing so, it is also important to protect the commercially sensitive package data since such data can be obtained by malicious entities, resulting in the theft of certain products from the ports.

In this paper, we address the bin-packing problem. We publish the data (such as  weight and volume) in a privacy-preserving way and use it for the optimization. For data anonymization, we propose two methods which are designed using differential privacy. Differential privacy (DP) provides strong privacy guarantees and pays attention to how much the existence of a particular row of data influences the final output.

1) Differential privacy with k-anonymity: We apply the exponential mechanism~\cite{mcsherry2007mechanism} to the mapping function of $k$-anonymity to achieve differential privacy.

2) Differential privacy with clustering: We first cluster the data based on the number of occurrences, and then add Laplace noise  to each cluster.

Blockchain demonstrator serious game

Serious game designed to demonstrate the potential impact of blockchain technologies on the supply chain in an online educative game based on the beer game.

Made by students and directed by Maxime Bouillon

A serious game to show what blockchain will change on our supply chains

The blockchain demonstrator is a game being developped for about a year now by Windesheim researchers in collaboration with teams of students. Designed as an educative game based on the beer game originally created in 1960 at the MIT, the blockchain demonstrator consist of four players set up in a traditional supply chain. They first go through a few rounds, facing the bullwhip effect before they are introduced to sets of new technologies that will potentially make it easier to organize their chain.

This game is currently under development, we have been prototyping and testing with groups of students, teachers and consortium members. When done, the game will be played in a workshop setup, with a facilitator/teacher and learning outcomes.

Other resources

The Spark! Living Lab is involved in several projects with various members of its consoritum. These are some of the reports of their work.

Made by teams of students and researchers