Ramvinojen Narayana Perumal

CAN, Automotive Ethernet communication and Vector Device HIL architecture design team

Gather, simulate data from a newly designed NMR sensor prototype. Design and validate different Deep learning models.

• Presented the Proof of concept to the management team, benchmarking 70 to 90% accuracy • Gather the data set from a newly designed NMR sensor, designed explorative models and validated Deep/Machine learning models to estimate complex parameters from the measured Fourier coefficients • Implemented CNN, Autoencoder, DNN and Bayesian hyperparameter optimization

• Development and programming of a tool for the evaluation, analysis and visualisation of correlation measurements on several vehicle powertrain test stands • Implementation of methods for statistical evaluation and outlier identification • Evaluation of test data with MATLAB and internal tools

Diagnostic and functional ownership of Global Central Gateway ECU • Created, tested and analysed the diagnostics for all ECUs of Daimler truck architecture. Also, supported the creation of “Hardware In Loop (HIL)” environment in Vector VT system for Engine ECU. • Co-authored a patent on the topic “Improving battery life of an autonomous vehicle by optimizing power consumption for inter-vehicle communication”.

Functional testing of Trucks Maintenance System Software in Central Gateway ECU • Created simulation environment and test script; executed the testing, analysed, drew conclusions from the results and suggested the necessary functional changes required.

Embedded systems, Machine learning, Deep Learning and Industrial Automation

Embedded systems, Control Systems, Process Control and Electrical and Electronic drives