Mayur Bokhani

Further developing the indoor‐navigation system created during the internship for autonomous drone flight testing, including test‐software development. Programming microcontrollers with PlatformIO and Arduino for low‐voltage electronics. Researching current projects, creating experimental setups, and collaborating with factory engineers. Preparing project concepts, overviews, and documentation using Jira, Confluence, Excel, and PowerPoint.

Thermal Anomaly Detection of Functional Elements in Industrial Facilities: A Dual‐Sensor Deep Learning Approach for Component Classification and Condition Monitoring (grade 1.0; written in English). Designed and manufactured a portable RGB and LWIR thermal‐camera module; built a ROS 2 data‐acquisition and communication architecture on Raspberry Pi 5; created and annotated a thermal dataset with synthetic anomaly generation; implemented and validated EfficientNet‐B0, ResNet‐18, and Deep SVDD models using Pyt

Conducted ideation and selected two drone‐use cases; defined the technical architecture and communication interfaces; designed and additively manufactured a drone frame. Built UWB‐based indoor localisation and implemented drone positioning in ROS 2 and Gazebo. Developed Time‐of‐Flight sensor simulations and EKF sensor fusion for UWB, IMU, and ToF data using C++, Python, ROS 2, Gazebo, and PX4.

Foundry internship: pattern design for moulds, moulding and casting of cast‐iron components, quality control, and final finishing for dispatch.