Graduation Assignment -  Failure detection in telecommunication network systems 

Are you interested in the practical applications of AI? Do you get enthusiastic about optimizing and implementing the right machine learning models for the issue you are presented with? Do you have the skills and ambition to deal with the complex mathematics behind prediction models, as well as to think along with the customer to define his problem? Then this graduate assignment at Bright Cape is really something for you! Within our multidisciplinary teams, we constantly work on optimizing processes by means of the latest data applications and techniques. In this way, we create value for both our customers and our internal products.

At Bright Cape, we strive to meet our client’s challenges across the full data science spectrum. Some of the topics our data scientists have been working on over the past year include:

  • Use of machine learning algorithms for predictive maintenance
  • Building forecasting models to predict the optimal price and stocks
  • Developing a control channel and an accompanying dashboard to provide insight into financial processes
  • Automatic customer screening by combining internal and external data sources
  • Process mining

Context of the assignment

The assignment is part of an anomaly detection use case aimed at maintenance and process control of telecommunication network systems. One of our clients is selling telecommunication network systems for 15 years of continuous operation. At this moment they have insufficient insight into when chipsets in these systems might break, resulting in reactive maintenance when it is too late. This is expensive and results in system outages and reputation loss. The network systems are continuously monitored on a variety of aspects, like chipset temperature and power consumption, resulting in a large amount of data that provides information on the functioning of the systems. The aim is to develop an anomaly detection solution that indicates when the operating behavior of chipsets deviates from the standard working conditions.

The assignment

You will have a scoped assignment within this larger project, aimed at doing research into potential anomaly detection algorithms and developing a pipeline that identifies potentially broken chipsets. The exact research question is to be determined in consultation.


  • Literature study to find a good solution direction for this anomaly detection problem
  • Implementation of the chosen solution
  • Validation of the implementation

Contact us!

Are you a motivated student with a strong mathematical background that is looking for a challenging assignment in the industry with a direct impact, do not hesitate to contact us! Apply today!

Contact person: Remco Theunissen

Do you have any questions about this graduation assignment? Get in touch.

Remco Theunissen

Challenge accepted? We would like to meet you!

Press 'apply' and share your motivation and resume. We will get in contact with you soon! 

Who we are?


Bright Cape is a  data consultancy firm located in Eindhoven & Amsterdam with a European reach. Extract, Embed, Educate is what we do. Extract value out of our customer’s data, Embed (Bright Cape proprietary) solutions into our customer’s processes & governance, and Educate them to get familiar with data analytics and our solutions.

Bottom line: We help companies increase their revenue, diminish their costs, and increase their process efficiency through Data Analytics & Science, Human-Centered Design, and Process Mining. Besides, we are constantly working on various astonishing Europe-wide innovation projects which have led to several innovative products.

Bright Cape employees are very diverse but share their excitement for new opportunities and a positive way of thinking. We also greatly value our quarterly team outings, monthly ‘VrijMiBo’s’, Christmas party, family day, and most of all our yearly ski trip.