Munich, Bavaria, Germany
On the topic of the master thesis - Analysis of methodologies for clustering in the field of connected vehicles on a cloud-based Big Data platform: • Completion of topic-specific training, among others to obtain the BMW-internal data driver's license. • Familiarization with Impact (i.e. Palantir Foundry), the BMW Data Portal, the Cloud Data Hub (CDH), and Amazon Web Services (AWS). • Conducting expert interviews with data scientists from different departments in the area of Big Data analytics. • Analysis of the current use of clustering methods and their potential in the BMW Aftersales environment. • Research literature on clustering methods for data, Big-Data platforms and create a summary. • Selection of the required data from the CDH and implementation of data pre-processing. • Implementation of different clustering methods using Spark and Python programming language in VSCodium integrated development environment. • Configuration and use of AWS with Amazon EMR, EC2, S3, SageMaker, and Glue services among others. • Evaluation of the implemented clustering methods and visualization of the results using Python and the libraries Seaborn and Matplotlib. • Creation of the master thesis based on the collected findings and results. • Regular coordination meetings.
On the topic of the bachelor thesis - Analysis of potential manipulation of vehicle mileage: • Research regarding hardware for the manipulation of vehicle mileage. • Evaluation of different detection methods regarding vehicle mileage manipulation. • Evaluation of different data channels regarding availability of vehicle data. • Data protection evaluation regarding vehicle data according to their processing in data models in coordination with the BMW Group Data Protection. • Conception of a Convolutional Neural Network, for the classification of vehicle data based on the LSC (Last-State-Calls). • Setup of a Python development environment. • Database queries in the Global Data Lake of the BMW Group. • Programming of a Convolutional Neural Network using the framework Tensorflow. • Programming of data processing and data management for data sets to facilitate the transfer to the data model. • Setting up a virtual client for training the data model. • Evaluation and presentation of the results. Working in the Mileage Verification agile team: • Participating in jour fixes and coordination meetings. • Adjusting the methodology of the bachelor thesis with the team. • Research regarding hardware for manipulation of vehicle mileage for ordering and installation in test vehicles. • Creation of first test scenarios for tests for valid data generation with manipulated department vehicles. Test and validation: • Organization and coordination of testing sessions with development vehicles. • Test case creation and documentation in HP Application Lifecycle Management for Mileage Verification. • Validation of the BMW Teleservices web interface in the context of regression tests.
Test and validation: • Test case definition and documentation in HP Application Lifecycle Management. • Conducting end-to-end tests of BMW Teleservices. • Further development of the test suite, including teleservice test application (Python programming language). • Organization and coordination of testing sessions with development vehicles. Data analysis and reporting: • Analysis of potential mileage manipulations. • Adaptation of the central reporting tool for the generation of regular market reports. • Analysis of incident tickets with a focus on online driver logbook. • Processing of teleservice calls relevant to data protection. Project management: • Alignment of the scheduling of departmental projects with the resource planning of the BMW Group IT interface partners. Further tasks: • Support in the internal departmental administration of test vehicles. • Inventory of departmental IT assets. • Editorial revision of the intranet presence of BMW Teleservices.