Omkar Prabhu Email & Phone Number

Thane, Mh, In

Thane, Maharashtra, India

- Designed and implemented a Bluetooth 5.0 communication framework utilizing Nordic SoCs and Zephyr APIs, facilitating seamless AWS-IoT integration via MQTT protocol. - Engineered robust security features like asymmetric encryption with multi-protocol sensor communication (UART, I2C), resulting in enhanced system protection. - Optimized memory management and thread communication through strategic implementation of Kconfigs and devicetree configurations. - Championed a cost-optimization project for a video doorbell product, successfully reducing BOM costs by 60% while preserving premium features and performance indicators like resolution, low light performance, encoding standards, power consumption, boot up speed.

Thane, Maharashtra, India

-- Managed a team of 4 engineers, facilitating knowledge sharing throughstructured sessions focused on system and code architecture.-- Drove the development of vital 5G RAN-AMF infrastructure, incorporatingessential authentication and session management callflows using protocols(NAS-NGAP over SCTP) using C++ and ASN libraries.-- Engineered RRC communication using multi-threaded architecture and JSON(nlohman JSON libs) encoding between UE and RAN-- Designed and implemented databases for UDR/UDSF components whileassessing performance metrics like CAAS, PAAS, development ease, scalability,and query timings across various database technologies.

Mumbai, Maharashtra, India

-- Engineered and refined 5G network functions (RAN-AMF-SMF-UPF) following3GPP standards and SDN framework, targeting workflows for registration, PDUsession, and service requests. -- Applied concurrent programming techniques with thread pooling to boost loadtesting efficiency. -- Executed designs in C/C++ utilizing various APIs (e.g., STL, ASN, Boost,Casablanca/REST) alongside Redis Database on Linux. -- Developed and implemented automated debugging tools for log and networkpacket analysis, significantly improving troubleshooting efficiency. -- Managed a CI/CD pipeline with Git and Jenkins, guaranteeing the deploymentof readable, reusable, and debuggable code.

Indore, Madhya Pradesh

As a budding startup we worked on these things1. Deciding problem statements to work on and do market research Searching for a problem statements mainly related to computer vision and machine vision, brainstormed on various ideas, gathered possible information about the bleeding edge research going on in this areas, problem statements other companies working on.2. Systems Design (Listing out machine vision cameras and lights) Went through reference manuals, studied different aspects of the cameras and light and research related to it. Created documents and spreadsheets based on this information, which will list out all sorts of cameras and their different aspects 3. Technical Worked on image processing and machine learning algorithms, for various projects like labels and packages inspections in logistics and supply chain. Currency inspection. Also, one project related to defense where we designed system which is camera based automated target scorer.4. Operations Listing out various company profiles, cost of cameras and lights, dealing with vendors to get best deals as possible

Boulder

Course Design and Overview This course teaches how computers really work, starting from digital logic. It covers basic CPU design built from combinational and sequential logic, an introduction to programming in assembly, and basic OS design requirements, with an emphasis on embedded system design. We use the Nios II assembly language on DE10-lite boards. Topics covered include register logic, number representations, basic pipeline processing, caches, assembly, calling conventions, memory layout, memory-mapped I/O, interrupts, and OS design.

Boulder

Goals of the project and brief description of each.Kindly checkout the research paper for more details:http://mercury.pr.erau.edu/~siewerts/extra/papers/XPO18-Drone-Net-Passive-Instrument-Network-Driven-MV-and-ML-for-UTM-Siewert.pdfAim: Drone-net aims at providing a low cast radar system to capture variety of drones which are not allowed in restricted zone like near airport1. Electro optical part along with machine learning So basically collect data from different wide angle view cameras placed strategically so that we can collect images of the drone and stitch them together so that we would know different aspects of it like what's it approaching speed, how much distance it is at using depth sensing techniques, and also use a database to create a profile for different drones and use machine learning to match these profiles in real-time.2. Acoustic partMy team is working on a part of this project that is direction of arrival detection of drone using sound signature of drone itself captured by MIC arrays. Currently there are few similar systems but they use RF technology which proves costlier and more power consuming. To detect drones we studied about frequency spectrum of drones , its variation with respect to time and space through analysis provided by various research papers and on that basis we have developed effective and efficient adaptive filtering method and we are using MUSIC ( MUltiple SIgnal Classification) algorithm to detect drones.Here we are using Mic array which will capture drone signature and that is interfaced to R-pi on which the MUSIC algorithm will run.MUSIC algorithm uses time difference between sound detection of the mics placed at different but known heights to calculate approaching speed

Master Of Science - Ms, Embedded Systems Engineering

Bachelor Of Engineering - Be, Electronics Engineering