Software Engineer II at Postman | Full Stack Developer | MERN | MEVN | Javascript

Pawan Kumar  • Peerlist

Artificial Intelligence based melanoma scanner. It can detect melanoma, a type of skin cancer, using snapshots of suspected area. It uses Machine Learning, Particularly Deep Learning and CNN to decide that the area is melanoma tumor or not. It is backed by 57,000 images of patients (lesions) that train the ML Model. It can achieve accuracy of 70%.

MelanomaScan

Technology Used: IOT,AVR Programming, Python and C++
Made an IOT Based Smart Goggles for Blind people which works on the principle of Ultrasonic reflections, which is used by bats to detect the nearby objects by bats. The person wearing the gadget gets 3D Heptic feedback on hand. That pattern is formed by data processed from multiple Ultrasonic transducers and recievers installed on the goggle by RaspberryPI. 3D feedbacks makes the person to sense objects in 3D within 4 Meters in range. This eliminates use of Sticks. We are currently working on this project and future version will include Voice Assistance, Google Maps (Audio) and Calling.

POLY Vison : IOT Based Wearable for Blinds

The idea was to develop a smart dustbin for the University which sends an SMS, when it gets filled, to the local garbage authority. It is an IOT based dustbin with various features required for smart city. A website was also developed which shows the location of the dustbins and the garbage level of the dustbins present in the University.
Technologies used: Arduino, C, HTML, CSS, JavaScript, PHP 


SMART BIN: Automatic Dumping and Statistics Monitoring 

This firmware helps to connect any device to Wi-Fi through serial port using ESP8266 board and can send or receive data in textual form. It is easy to program and use. One can connect Arduino board to any server using Wi-Fi by just writing 4 lines of code. It applies interpreter-based approach for accepting commands. It is based on Wi-Fi library provided by ESP8266 community package. It can connect to any server or website to transact API based data. It supports SSL encryption, basic authentication, ping, traceroute and several others.
Technologies used: C++

Custom Firmware for ESP8266 Board

Made an application in react native for student organization cell at Lovely Professional University. It supports push notifications, events, accounting, e ticketing and attendance management system.

LPU SOC App

Driver-less Solar Bus : A bigger, smarter move towards new India.
1 Year long, Multidisciplinary, Innovative project exhibited in 106th Indian Science Congress

Developed by me with love and affections for Honorable PM Narendra Modi.

Bus Inaugurated by HRD Minister of Govt. of India, Prakash Javadekar

The bus has following abilities :
1. SAE Level 2 Driverless Capabilities
2. Maximum Speed : 40KM/h
3. Solar Charging / Electric Station Compactible
4. Seating Capacity :14 (4 in this model especially for Guests at 106th Indian Science Congress ) 
5. Upto 70Km distance with a single charge.

Technologies Involved in Driver-less technology:
Lane Detection using Image Processing and Magnetic feedback.
GPS Grid based Navigation
3D Obstacle detection using Lidar Lite and High range ultrasonic sensors
Self learning using deep learning for situations of signal loss
Inertial sensors for state detection

Solar Driverless Bus - LPU

A Single Page Wallet Application based on Object Oriented PHP 7 Design and React, React Router and Redux Based Application

Features of Wallet Application
 Send and Receive Money
 Single Page Application with offline support using service worker (to be enabled in index.js)
 Can serve as payment gateway using blind payment address (payment api urls)
 Schedule future transactions
 Loan Request and Approve (through Admin)
 Multi User Level Architecture (Admins and Users)
 Mobile Friendly Responsive Architecture
 Gift Cards issue and redeem
 Dual Entry Accounting Methodology
 Login, Register and Profile Update Support

MegaWalletX

AI based Mid-Day Meal forgery detector, made during Smart India Hackathon
- Both cameras send image to server where 5 different ML models compute the dish cooked, number of people served, meal attendance of students, food quantity prediction (reduces food wastage), quality detector.
Technologies Associated: Python3, Node.js, React.js, OpenCV, Cisco Meraki API, Tensorflow, React Native, MongoDB

Mid Day Meal Monitor

The Organization Management System is a product level project which was designed for all the student organizations inside Lovely Professional University. It was designed using popular PHP Framework Laravel and Vue.js. It had various features like enrollment of students for events through coordinators, peer to peer transactions, custom authentication, authentication using queues, self-registrations, online meetings, online attendance, smart card QR based delocalized attendance and enrollments, budgets, accounts, reports, gift vouchers, Payment gateway etc.
Technologies used: Laravel, Vue.js, HTML, CSS, JavaScript
visit https://demo.nuoms.nukrip.com with credentials
username: 11711106
password: demodemo
