Mathematician and Data Scientist with experience as a Junior Data scientist at DataBI and as Head of electronics department at Formula Student team UnexMotorsport. Has worked on PowerBI projects, process automation, soft…
Mathematician and Data Scientist with experience as a Junior Data scientist at DataBI and as Head of electronics department at Formula Student team UnexMotorsport. Has worked on PowerBI projects, process automation, software development, telemetry, and data science, and holds degrees in Mathematics, Secondary Education Teacher Training, and Big Data and Data Science.
<p>Plan tasks following actual regulations<br>
Assign tasks following actual regulations<br>
Supervise tasks following actual regulations<br>
IT specialist<br>
Software development<br>
Telemetry<br>
Data science</p>
Plan, assign and supervise tasks following actual regulations. IT specialist in software development, telemetry and data science
<p>Developed different projects including PowerBI<br>
Automating processes</p>
<p>Prepare necessary material for conference members<br>
Organise different lectures along the conference</p>
Prepare necessary material for conference members and organize different lectures along the conference
Outlier Detection in Racing Car Telemetry Data as a Fault Diagnosis Tool<p>The present Master’s Thesis aims to develop a methodology for anomaly detection in motorsport telemetry data, with the objective of identifying abnormal behaviours that may be associated</p><p>with mechanical failures. Based on real data recorded by the acquisition system of the Ligier JS2R</p><p>vehicle, various physical signals related to the engine, gearbox and driving behaviour have been</p><p>analysed using unsupervised machine learning techniques.</p><p>The study focuses on the analysis of a real failure that occurred during a race — the explosion</p><p>of the gearbox oil radiator. Since there were no labels indicating the exact moment of failure, an unsupervised learning approach was adopted. Furthermore, due to the temporal nature of telemetry</p><p>data, the project relies on multivariate time-series techniques, which make it possible to model</p><p>the normal behaviour of the vehicle and detect significant deviations. To this end, four anomaly</p><p>detection methods were implemented and compared: Hierarchical Temporal Memory, Matrix Profile, Unsupervised Anomaly Detection, and Anomaly Detection with Correlation</p><p>Analysis.</p><p>The results confirm the ability of the models to identify the moments in which the behaviour</p><p>of the variables deviates from normal conditions, highlighting gearbox oil pressure (gsp) as the</p><p>variable most strongly associated with the failure. In addition, the alterations observed in engineand driver-related variables reveal the dynamic consequences of the incident.</p><p>Finally, this work lays the foundation for the future development of an interactive tool that</p><p>will allow engineers to visualise detected anomalies along the vehicle’s track or by laps, thereby</p><p>facilitating post-session diagnostics. Despite certain limitations — such as the absence of spatial</p><p>information and the need to expand the database of failure cases to obtain a more comprehensive</p><p>tool - the proposed methodology demonstrates strong potential to become a support system for</p><p>data analysis in competitive racing environments.</p>
نظام التوظيف التشغيلي المعتمد على الذكاء الاصطناعي أولاً، الذي يوحّد ويؤتمت دورة التوظيف بالكامل للمؤسسات الحديثة حول العالم.
د.جوب منطقة حرة ذ.م.م. 2026 © جميع الحقوق محفوظة
