Sanal Darshid Ramkumar

Winners of Hinterland Hack 2020 - Piening Challenge Hinterland of Things Sep 2020 Company: Piening Personel GmbHDepartment: Piening Pioneers – The digitale Task Force of the Piening GruppeProblem: The entire recruitment process is not sufficiently efficient and at the same time is perceived as a negative experience by most potential candidates.Challenge Description: To find a perfect and sustainable matching algorithm that is backed up by a personal applicant experience. To give revolution of the application process.

Finalist of “City Hack 2020” Hackathon - May 2020 Quality Health Challenge:Development of Personal Assistant for the Health care industryDeveloped a Chatbot for Personal Assistance

Winners of RoboCup 2014 Vircent Technologies Pvt Ltd , Microsoft Apr 2014 A National Level event on Robo Race for Line follower and obstacle avoidance conducted by Indian Institute of Technology(IIT) Kharagpur

Finalist of Ultra Hack Hamburg 2019 Hackathon - Traffic Management System Ultra Hack 2019 Challenge:As soon as infrastructural projects are executed, the overall system of Hamburg Harbour is put under stress. With the Hamburg Harbour being located in the heart of the city of Hamburg, many logistical challenges arise with regard to traffic in and around the harbour, whether it being shipping, road or rail traffic. A functioning and efficient slot management for trucks are as essential as state-of-the-art software solutions for transport management, customs clearance or veterinary… Show more Challenge:As soon as infrastructural projects are executed, the overall system of Hamburg Harbour is put under stress. With the Hamburg Harbour being located in the heart of the city of Hamburg, many logistical challenges arise with regard to traffic in and around the harbour, whether it being shipping, road or rail traffic. A functioning and efficient slot management for trucks are as essential as state-of-the-art software solutions for transport management, customs clearance or veterinary checks.Solution:A traffic recognition system using deep learningDescription:• The collected traffic dataset was classified as dense traffic, accident, fire and sparse traffic• State-of-the-art convolutional architecture ResNet-50 was trained and tested with the labelled dataset• Using Unreal Engine, Carla and Vector Zero a virtual map of Autobahn A7 was created for testing the prototype• The model was trained using Nvidia GeForce GTX Titan X and deployed on Raspberry pi 4 B mounted on top of a small prototype resembling a car or truck• The autonomous bot detects the density variation (i.e Dense, Sparse ) in and around the harbour• Based on the traffic density it can re-route to the destination with next possible shortest path, thus reducing the traffic jam and time of arrival• A prioritisation algorithm based on predictive analysis with the provided dataset using Wi-Fi direct • Communication was established between the autonomous bots Show less