On 3rd May 2019, I had a wonderful opportunity to have been invited as an IEEE SA (Institute of Electrical & Electronics Engineers Standards Association) Jury to evaluate an IEEE innovation competition named “Prayojaka” that was organized by the IEEE Student Branch of ASIET (Adi Shankara Institute of Engineering & Technology).
The event started at 12:30 PM and all the students & their groups were ready on their respective exhibit areas to show their project work. I was asked to evaluate the Computer Science projects along with 2 other judges. The 2 other judges were Computer Science professors from 2 famous engineering colleges. I represented the jury panel as an industry expert. My wife “Sakshi Mehta” (Quality Assurance Engineer – Suyati Technologies) also accompanied me for the event representing Suyati Technologies to assess the quality of work & the depth of thought put by students on their projects.
We evaluated 6 computer science projects.
- Using Machine Learning to generate HTML from images
- Smart Transportation using IoT & Big Data Analytics
- Smart Vaccum Cleaner
- Asset Tracking using Cloud IoT
- Android Theremin
- Home Price Estimation using Predictive Analytics
All the 6 projects were stunning concepts & the level of hard work that the students had put in for this competition was much above my expectation levels. All projects were done using Python & Machine learning libraries.
Using Machine Learning to generate HTML from images
The concept of this project was to generate HTML pages from wireframes or sketches. We all know that a considerable amount of time spent in the SDLC is during conceptual design. This is the phase where we draw wireframes & mockups to illustrate how we envision the system. Later, once the wireframes are approved, the UI designer creates the User Interfaces based on the wireframes. The idea of this project was to reduce the time spent on creating UI elements from scratch by referring to the wireframes/sketches. The student had developed a python neural network program that had learned different shapes and also text identification models with which the program would create a normal input element of type=text for username and an input element of type=password for password fields. The program was able to identify different shapes of buttons and tables and create HTML pages within seconds. Yes, there is always a point of CSS styling (that he plans to work in future) which was not yet programmed at the time of project demo.
Smart Transportation using IoT & Big Data Analytics
The concept of this project was to predict roadblock on a map with respect to time. The team used google maps data of a few months to train their prediction model and plotted a heat map sort of visualization on a google map as you slide the time to a future date/time. The amount of work related to training the model for accuracy & precision was indeed worth appreciating. With respect to the relevance of the project, some industry solutions already did exist. Their vision was to develop this framework first & then relate this prediction framework for industry verticals like hospitals to know their efficient means of transport in advance.
Smart Vaccum Cleaner
The team had created a prototype of a vacuum cleaner that can automatically detect obstacles and clean the room both at floor level & at ceiling level. The vacuum cleaner was designed using Arduino and had an electronic lifting mechanism with which it can climb through the wall and reach the ceiling. The machine had a chamber to collect dust which can be manually opened & emptied. The prototype, unfortunately, functioned improperly during the project demo. One aspect that stopped them from creating a robust machine was budget. Most of the electronic chips were costly and the team members couldn’t afford to spend much from their pockets. Some suggestions from the jury were to think of incorporating Python + Arduino interfacing with which they get the full power of Python programming and they don’t just have to limit themselves to Arduino programming capabilities. The lifting mechanism was based on vacuum cleaner being hoisted to the ceiling via a string cord which we felt wouldn’t be practical enough to help people use this machine. A suction based prototype that can fly and stick to the ceiling based on the suction force and in turn vacuum clean the roof would be ideal.
Asset Tracking using Cloud IoT
The idea of this project was to track humans (or what he refers to as Asset) entering a campus/workplace so that they don’t have to punch-in and punch-out every time. This project would solve lots of problems around the misuse of access cards, improper access records leading to salary calculation errors, people having to wait in queue if a huge crowd has to enter the campus or an event hall, etc. The business problem was very unique and had relevance. The team had chosen WIFI routers installed in the campus as an event source. How this works is, when a student enters an area with his mobile phone would automatically get registered in the WIFI zone. The huge work that the team had put in was to map the entire college campus into small WIFI zones and feed this mapping into a tracking program written in Python. The team chose WIFI routers for cost reasons instead of going with RFID chips & IR detectors. This project was more of computing loads of data captured from access points to understand where an asset is currently at. The use case is highly relevant in the campus to track students, factory to track assets & individuals from approaching no access zone, etc
A theremin is an electronic musical instrument controlled without physical contact by the thereminist. This was patented back in 1928. When the thereminist moves their hands closer to the antenna, the pitch of sound changes that are emitted by the instrument. With careful movements & good practice, a thereminist can produce music using hand movements. The goal of this project was to develop the theremin in Android as an app so that anyone can use screen touch movement + spatial gyroscopic movements to create a melody. A great technology of the olden times was nice to be seen in a new revolutionized way in Android. The use-cases of this project are still unclear as many other sophisticated musical instruments & synthesizers exist. However, a musician from his heart can create music just with his phone rather than investing in a costly musical instrument was a good takeaway.
Home Price Estimation using Predictive Analytics
The team displayed a stunning predictive model that used apartment & villa data from many websites & created an Estimation portal. This portal would ask few questions related to what the end user is looking out for (like Apartment/Villa, how many bedrooms, area expectations, facility expectations, etc) and choose a locality + the year in which they plan to invest. Based on these inputs, the predictive model would display the estimate of the price of the home at the chosen year with the selected amenities. This concept was really unique and we felt that an average buyer would definitely look for these kinds of predictions so that he/she can plan his savings & investment wisely. The entire program was developed in Python with Regression based Predictive Model that looked at patterns of home prices & predicted prices at a future timescale.
The jury had a wonderful time interacting with the teams & exchanged lots of ideas as to how they could improve their project relevance from an industry standpoint. We helped them understand how they could do the model comparison, select of models based on accuracy & precision, thinking of practicality, the impact of the solution to touch human lives, etc that bring the unique flavor to an innovative project. Yes, the budget was a constraint for most of the teams but I guess IEEE, alumni associations and even interested organization could easily solve these issues for our young innovators to be more successful with their project prototypes.
Well, the entire competition ended at 2:30 PM. All the student teams and jury were fully exhausted by then.
Shortly, we announced the winners of the competition and the IEEE IAS Chapter Advisor handed over the jury certificates to all the Jury Members. I thank the IEEE SA and IEEE Student Branch at ASIET college to have invited me for this technological extravaganza. I also thank Revathi Krishna(Chief Communications Officer – Suyati Technologies) to have taken interest in allowing representations from Suyati family to mentor our young innovators. It was definitely time worth spent. I hope to have such interactions in the future with you all once again.