Ambarish G Kaliamurthi

- Working with Unitree Humanoids G1, Quadruped dog Go2 and Robotics Arms - Teleoperations and Data collection

- Developed real-time indoor navigation using ROS2, Nvidia Isaac Sim, Jetson Nano orin, Nvidia Isaac ROS, and teleoperation frameworks, incorporating advanced motion and path planning. - Built a sensor fusion research pipeline combining LiDAR, depth cameras, and MMDetection3D for 3D object detection and obstacle-aware navigation. - Explored collaborative compute frameworks with ROS2 nodes to ensure robust, low-latency communication between onboard systems and remote interfaces. Two current research: 1. MoRAL: A 3D reasoning framework for ADAS systems 2. TheRobotSim: A unified hardware aware orchestration tool for robotics simulation for deployment efficieny and quality assurance

Automation - Experience integrating third-party IoT devices with Edge. - Developing, piloting, and delivering demos, product overviews, and architecture to customers - Managing and maintaining services Concurrency - Improving the efficiency of I/O bound calls by using async (asyncio) to make constant time from linear time - Multithreading device upgrades to reduce the time from linear time by a factor of five Edge - Brainstorming cyber-physical system designs with the senior team to solve complex client problems - Developing Edge SDKs for client usage - Finding vulnerabilities and reporting them to clients Experience working with a range of device types, including cameras(AXIS, Hanwha, ONVIF etc), NVRs(Ambient, Exacqvision, milestone etc), occupancy and light sensors, AV systems, IIoT devices, alarms, and NUCs etc.

Developed a SaaS application from scratch called Byonic doing end to end work and was praised by the customers during PoC Web Application - Creating a web app with SQL CRUD operations in the B2B martech space - Implementing REST APIs for handling large volumes of data (over 10 million records) - Migrating from SQL to an object-relational mapping (ORM) framework - Deploying code in the UAT environment without downtime Agile environment - Collaborating with the Scrum master and developers to provide clear design specifications - Using pair programming to ensure high-quality code Security - Implementing a secure user authentication and authorization system using one-way hashing and jwt tokens for multi-tenant operations. Email micro-service - Implemented SMTP micro-service to automate system-generated emails for users. Experience with a range of technologies, including Python, flask, Typescript, Angular, MySQL, sqlalchemy, nginx, mongodb, scikit-learn, pandas, POSTMAN, SoapUI, git, JIRA, and Confluence.

Edge-based solutions with ARM Cortex - Implementing prototype edge solutions on ARM Cortex for a smart kitchen appliance. - Includes fetching and processing data from connected things, such as sensors and devices. AWS IoT Greengrass - Used AWS Greengrass to bring AWS functionality to edge devices, allowing them to act locally on the data they generate, while still using the cloud for management, analytics, and durable storage. - Implemented firmware over-the-air (FOTA) updates using the MQTT protocol for the core device. - Used systemd to create service to automate the process AWS Lambda for Edge devices - Implemented Local Lambda functions that can be triggered by local events, messages from the cloud, and other sources, which brings local compute functionality to connected devices. - Used Lambda functions to filter device data before transmitting the data to the cloud. - Optimized the Lambda functions to improve their performance and avoid errors. Automation - Built in-house solutions that brought down setting-up core from 1.5 days to 1 hour, used by the whole organization.

◦ Mall-bot assistant: Built a voice-based software component for their MallKiosk, which uses Google’s Speech Recognition API for the audio transcriptions (STT), Intent identification using patterns acting as Spoken Language Understanding, Text-To-Speech (TTS) using pyttsx3 and gTTS for converting text back to speech for responding. ◦ Anti-Social robot: Implemented Social distancing robot using Keras (TensorFlow) pretrained YOLOv3 and OpenCV for their realtime community video streaming service on an embedded device that can be controlled remotely. ◦ Mask detection: Implemented Face Mask Detection using Keras (TensorFlow) ImageNet pretrained MobileNetv2 and OpenCV, finetuned on a proprietary dataset for their realtime community video streaming service on an embedded device. ◦ Coursework creation: Created 30 hours coursework on Robotics, Artificial intelligence, and Game development for undergraduate STEM students. ◦ Mentor: Mentored undergraduate students on AI and Embedded systems. Also taught primary and secondary students robotics.

◦ Self-driving Industry robot: Worked with a prototype self-driving robot for industrial cargo movement. Planned paths along with maps and detected lanes with ZED stereo camera and Picam with Nvidia Jetson nano, using ROS and Rviz. ◦ DataMatrix Detection: Wrote code to detect Datamatrix using pylibdmtx for asset tracking, identification, and data-driven applications of the objects going on Conveyor Belts in TVS warehouses sites. pzbar library was also included to detect QR code and bar code for consumer cargo. ◦ OCR on metal engravings:: Built Custom Deep Learning-Based end-to-end OCR model with attention. Attention Ocr TensorFlow with the FSNS dataset was tested. Tesseract OCR engine, Labeler was used to identify the OCR. The model was built in Nvidia Jetson Nano. ◦ Studied Reinforcement Learning for robot transfer learning:: Reinforcement learning of the robotic arms by the OpenAI were studied and test with the gym package. Simulated goal-based tasks for the Fetch and Shadow Hand robots. SLAM algorithms were studied to implement with self-driving cars.

Coursework: Data structures and algorithms in C++, Artificial intelligence and data engineering, Microprocessor Design.

I was highly involved in the CSI, Robotics club and IETE, where I had the opportunity to work with a diverse group of individuals to plan and execute various events and projects. I learned the importance of effective communication, organization, and problem-solving. I had to work closely with my team members to delegate tasks, set goals, and ensure that everything was running smoothly. I also had to be adaptable and flexible, as I often had to deal with unexpected challenges and setbacks. But perhaps the most valuable lesson I learned from my college experience was the importance of teamwork. I learned that by working together and supporting one another, we were able to achieve much more than we could have on our own. I also learned the value of collaboration and the power of diverse perspectives and ideas. Overall, my college experience taught me a wide range of skills that I know will be invaluable in my future career.

Passed Grade 12 in CBSE (Central Board of Secondary Education) with 80% score and distinction.