Software Engineering Competition in Suceava - Best Innovative Minds 2021

Annually, the software engineers working at ASSIST Software have a fantastic opportunity to flex their creative and problem-solving muscles by participating in a company-run software development competition.   

The innovation competition for software developers, designers, and students brings the most thought-provoking and innovative ideas and pairs them with out-of-the-box implementation techniques.   

The ASSIST Software event lasts almost half a year. Many have their eyes on it because projects initially presented at BIM generally go on to be used commercially with national and international recognition. 

Best Innovative Minds is considered a tradition amongst ASSIST Software employees. Everybody has a chance to win the grand prize of 1,200 euros and enter the Best Innovative Minds Hall of Fame. Add to that the possibility of winning other smaller cash prizes for adjacent award categories, and you have an event you can’t say no to! The total prize pool for this year’s event was 4,500 euros. 

At ASSIST Software, we encourage the development of technical and personal skills and Best Innovative Minds creates the perfect context for such growth. Collaboration and team spirit are also hugely important, since those who submit ideas to the competition must work with other teammates of their choosing in 100% self-managed projects.   

Not only is Best Innovative Minds a fantastic way of increasing technical knowledge, it also helps talented students from the Faculty of Electrical Engineering and Computer Science at the Stefan cel Mare University get a taste of what working at a software development company entails through our student selection process.   

In 2021, the Romanian software engineering competition started on the 22nd of July. All ASSIST Software engineers were invited to register and give a brief and unique presentation of their innovative software idea. This was known as ”The Selection Stage”.   

We are constantly looking for ways to improve and tweak the competition to keep it interesting and fresh. For example, in 2020, an amendment to the rules introduced a new award category - “Most Disruptive Technology” - for the project that used a less popular or recent technology. However, this year’s rules modification upped the challenge even more.   

In BIM 2021, contestants were allowed to compete by using one of three organizer-suggested ideas. All suggestions included big twists, such as making a project come to life while “writing as few lines of code as possible”.  

Furthermore, to reward the courage of those who took up the challenge of entering their ideas in the competition, a new raffle was added as well. Idea-submitters were given the chance to win a two-person all-inclusive weekend getaway somewhere in the beautiful region of Bucovina.   

After many deliberations, a small jury selected the seven most promising tech and software projects to pass to the implementation stage of the innovation competition. Student interviews and selections also happened at this time, since each team had the opportunity to choose the student they would be working with.  

The implementation stage lasted less than two months. Thus, the teams faced the challenge of creating a demo by the 4th of December when the Grand Final was held at the ASSIST Software headquarters in Suceava, Romania.   

All the projects presented at the Best Innovative Minds competition showcased the dedication, skillfulness, and perseverance of the software engineers that worked extra hours to deliver a top-notch software solution. 

-> Team members: Andrei U., Andrei V., Alexandra J., Cristi V.  

Behind the clever name, NFTeams  is a project that proposes a gamification solution to employee engagement issues many companies face today.  

According to Achievers Workforce Institute's fourth annual “Engagement and Retention Report”, only 21% of employees are “very engaged” at work. With disengagement costing up to 550bn US dollars a year, it is not surprising that organizations would greatly benefit from this innovative software where each employee gets a character minted (NFT) when they join a company and is motivated to then improve their avatar by completing challenges and missions.   

Taking part in adventures (company events) gains the avatar virtual currency. Tokens can then be exchanged at a marketplace for visual items of various scarcity levels, extrinsic rewards, or even fun prizes such as throwing a pie in a line manager's face or getting them to buy good coffee. 

The solution from NFTeams is a Web3 app connected to Polygon blockchain’s main net. Two smart contracts in Solidity define the core actions (for the internal token and the NFTs), and these are hosted on Infura via an API. Company assets (badges & avatars) are hosted on IPFS using a service called Pinata. This allows for the"minting" of ERC-721 NFTs, where the metadata of an item is stored in a specific wallet, thus creating ownership. The wallets are created automatically upon sign-up, and we use real MATIC for the fees. It only costs $10 for 5000 transactions. 

Features developed: 

• Gamification items (characters, achievements, leaderboards, stats, quests) 
• Reward system 
• MS Teams bot 
• Social platform features 
• Smart contracts 
• NFT minting and storing 
• Digital wallet 
• Internal token 
• Admin center 
• Main net connectivity 

What are the advantages of NFT gamification?  

• Increases user engagement - gamifies daily routine, making it more fun  
• Builds brand awareness - sharing externally and gaining appreciation  
• Encourages brand loyalty - fosters a sense of community  
• Improves motivation - creates a connection and a sense of ownership  
• Allows you to remain competitive - in the new, tokenized world  
• Eases the work of the marketing department - through social seeding performed by employees 

-> Team members: Petru C., Sergiu L., Bianca B., Marius P., Sergiu M.  

The “Democratie de colectie” project name roughly translates to “collectible democracy”, and it will all make sense once you understand how it works.  

In 2020, Romania ranked 62nd on the global index for functional democracy. This is the country's lowest score since investing in education about democracy. Therefore, this application tries to shed some light on the work activities of Romanian MPs. By accessing the data on the platform, electors can have a clearer perspective of these activities and can choose to compare two representatives or two parties.  

Considering that the official government-run website for the parliament contains insufficient data, the team made a demo of a citizen-friendly platform that included the following features:  

• Scrapes Romanian representatives' data from the https://parlament.ro public platform. 
• Saves the scraped data to a BigchainDB blockchain node. 
• Aggregates and displays data based on activities, and groups data by party, county, etc. 
• A notification system sends a monthly report for all the activities of the representatives from a specific county. 

The flow starts with the web scraper which runs once a month to collect  data on the activities of each representative in the past month from the parlament.ro official platform. Once the data is collected and transformed into a JSON ready format, it is stored into a PostgreSQL database and part of it, the unique data, in the BigchainDB node. 

Next, the aggregated data is displayed on the platform and an email is sent to people that are interested in finding out more about the representatives from a certain county in the last month. 

In addition to the classic API-based app with the frontend decoupled, we added a blockchain-like functionality by using BlockchainDB, a database with a blockchain characteristics library. Each of the microservices used to build the app were dockerized to achieve a quick and smooth deployment.  

What are the advantages of such a platform?  

• Transforms quantitative data into appealing and trustworthy visual pieces of information  
• Gives a new perspective on each candidate, come voting season  
• It’s a democratic exercise for citizens based on facts.  
• Stimulates engagement and makes information accessible to everybody 

-> Team members: Razvan P., Alin P., Andrei V., Cristian I.  

CopAI is a project that confidently tackles a worldwide problem that is causing financial losses every day. Rush hour traffic is one of the concerns of many city councils all over the world. Some cities do not have the necessary budget and architecture to change and bring urbanism plans to another level.  

The CopAI team’s main statement is that a single traffic light will not ease a city's traffic, and that it is necessary to develop an entire decision structure that knows which streets to clear and when and how to let pedestrians pass.  

The project set out to use two types of machine learning systems. One recognized cars, pedestrians, and other traffic participants, and the other learned about the traffic and made better decisions daily on how to ease it. 

Developed features:  

• A Unity simulation of a city (Suceava) 
• An ML system that recognizes cars 
• An ML system that  analyzes data from images and calculates new timings

What are the advantages of CopAI?  

• A faster city is more tailored for investors and companies  
• It can save residents and tourists' time  
• Car pollution will decrease as a direct result of the traffic decrease. 

-> Team members: Ionut M., Tudor A., Paul V.  

Another traffic-centered project, came about due to the fact that Romania ranked number one in the EU on the frequency of mortal car accidents in 2018.  

The project’s features include a deep learning system as well as a car crash alert, passenger monitoring, night vision, and a mobile app that analyzes one’s driving style. Therefore, Drive Vision could potentially transform into the best driving partner. It not only looks out for the driver and passengers but also tries to develop preventive driving skills.   

Technologies such as PHP, Python, C++, Kotlin, Pusher, and Google Maps JS were used for the demo and a working prototype that can be installed in any type of vehicle that uses an OBDII interface is certainly a challenge.  The team managed to create a Python library that can make the Raspberry Pi module communicate with  a vehicle in real-time to be able to log all the data necessary to generate reports at the end of  a driving session. 

The Drive Vision product can be used by private fleets to track  people that  are renting or  working on a vehicle provided by a company. It can be implemented by car insurance companies that want to create a profile for each of their drivers. The project can also be public and help  individual drivers improve their skills by using the reports generated by the mobile application (about fuel economy, focus, and behavior). 

Features developed 

• Passenger count detector 
• Driver monitoring 
• Emergency system / Crash alert system 
• Alert system for driver 
• Night vision (analyzes all the obstacles in  real time) 
• Mobile application ( analyzes the driving session) 
• GPS monitoring (real-time analyses of the current driving session) 

What are the advantages of Drive Vision?  

• Allows the building of a system that can be embedded in  a car  
• Reduces the number of deaths on roads due to the co-pilot aid  
• Alerting the authorities and passersby of a car crash enables help to get to the location more easily.  
• Encourages preventive driving 

Team Learnity  challenged us to upgrade to a new way of learning by exchanging the traditional chalk and board in favor of a ”limitless” possibilities VR platform. The demo highlighted new and fun ways of learning geometry, astronomy, and biology.  

This product can be used in class when teachers want to present a specific subject that students could then analyze from all angles. As an example, the teacher and the students could be in a “scene” surrounded by geometric shapes, and they could make geometric transformations to them (such as translations, rotations and scaling), by using hand gestures. It could also be a great way of teaching about the solar system; Students could  explore each of the planets and “slice”  them to see the inner layers.  

The Learnity project demo included three such subjects of study or “scenes” in the app: the human muscular system, geometric figures, and the solar system. The HUD shows the user all the gestures that are available, but it can be removed/shown by a simple finger snap sound. 

A ”pinch” gesture makes a whiteboard appear and the user can write on it while holding the same pinch gesture.  

To achieve the best user experience, the team increased the FPS (Frames Per Second) as much as possible. The app renders a lot of graphics in real time, and it is expensive for CPUs and GPUs. Therefore, the number of frames is filtered. In the “view object” mode, the app checks each 3rd frame. However, if the app is in “drawing on whiteboard” mode, it then checks each one.  

Features developed:  

• A platform for creating customized scenes 
• A platform for sharing custom scenes 
• Shared scene sessions (e.g. teacher + students) with voice chat 
• Add interface On Air to enhance user experience - Users can ”touch” buttons to press them. 
• Optimization on rendering  
• Physics experiments 

What are the advantages of Learnity?  

• Creates an add-on application for 3D figures using the LiDAR scanner (available for iOS)  
• Creates add-ons for personalized gestures that control specific 3D objects  
• Ability to share learning files (3D models + sets of hand gestures to control objects) 

-> Team members: Alexandru S., Alexandru C., Alexandru T., Iura G.  

With Higher Protocol, blockchain made another appearance in the series of presentations.  

 According to a survey by CareerBuilder, 75% of HR managers have found at least one lie on a CV and nearly 20% of hiring managers spend less than 30 seconds looking at each one. These factors make it impossible to know how many fabrications go undetected. In addition, some employers pour many hours into the lengthy process of verifying the information on CVs they receive. Therefore, the Higher Protocol team wanted to create a solution to address all these issues. 

Their proposal revolves around using blockchain to check backgrounds and employment history. Users would create digital wallets that, instead of holding cryptocurrencies,  would hold their personal information. To be visible in the blockchain, this data  would have to be validated by other users (companies, universities, academies, etc.). 

Higher Protocol assures that employee data would be encrypted and immutably recorded. This way, applicants would need to provide potential employers with permission to access their blockchain-based employment data.   
 

Features developed: 

• Background and employment history checks on blockchain - employee data encrypted and immutably recorded 
• Smart contracts - Incentivize seniors to complete interview tasks by releasing rewards once the tasks are done. 
• Applicants  can provide potential employers with permission to access their blockchain-based employment data. 
• Video assessments with A.I. sentiment analysis for a better matching score. 

In terms of technology, Higher Protocol used Heroku to create the test servers. They created two servers: one for the backend and one for the frontend. To create the frontend, they used NextJS because it comes with powerful server-side rendering (SSR), and static site generation (SSG). 

The backend was created using Ruby on Rails. For the blockchain, they used Ethereum, Truffle, Ruby ETH, and BigChain DB. The web app database was created on PostgresDB. 

What are the advantages of Higher Protocol?  

• Automation of the everyday, routine activities of a human resource manager  
• On a blockchain, records can be encrypted and immutably recorded.  
• Information can be verified by other users.  
• When an employee learns a new subject or completes a course, it will be made visible in the form of a block. This creates a verifiable and correct history of training activities and certifications.  
• By applying analytics to the data, companies would be able to match individuals to roles much more accurately and effectively. 

-> Team members: Dumitru M., Dragos T., Diana B.  

Cleaner Nation is an iOS and web application tailored for Romania that aims to unite people in the fight against littering and improper garbage dumping. However, the app can also be used by citizens in every country in the world.  

The Cleaner Nation team created an independent system and used machine learning for waste validation. 

This BIM project targets people who want to do good by cleaning or by adding places that need cleaning in their area. Utilizing the app, a user obtains a score that could bring them a monthly prize. The app was designed to be a way of having fun with friends or other registered users from the same area. 

The mobile app uses machine learning and a camera to detect if  a photo contains waste or not. People will see waste points on the map in real time. 

Features developed: 

• Register by phone (with country selection) 
• Show all waste points on the map 
• Route to waste point 
• Report waste point with ML validation and user location 
• Check the user rating 
• Check the user’s approved, done, and in-progress cleans 
•  Approval of waste points and clean points by other users 
• Show contacts that have the app and invite others 
• Share waste points 

What are the advantages of Cleaner Nation?  

• Solving a problem at a relatively low price (money for the monthly top  three citizens from each location)  
• Educating the future generation to take care of  their communities by gamification  
• Bringing the community together 

After discussing each demo and asking the right questions, the nine jury members decided who deserved to win the software engineering competition.   

• 1st place winner: NFTeams - 1200€   
• 2nd place winner: Democratie de colectie - 900€  
• 3rd place winner: CopAI - 650€  

SPECIAL AWARDS:  

• Most Disruptive Technology: Learnity - 500€  
• ASSIST Innovation Lab: NFTeams - 500€  
• Popularity Award: Drive Vision - 350€  

Worth noting is that the jury consisted of experienced software engineers working at ASSIST Software and university professors who kindly agreed to lend their abilities yet again and help us decide on the best projects. Their input was much appreciated, and their questions allowed us to figure out what teams should be declared the winners.   

Even though the 2021 edition of Best Innovative Minds finished not long ago, we are already excited for the next one. Through such challenges, we grow together and expand our pools of knowledge. BIM helps all participants develop soft skills such as leadership, communication, and time management, but it  is also the catalyst for unending discussions about technologies, workflows, and implementation processes.  

We are confident that some of these ideas will be expanded and will continue to receive peer praise and recognition.  

ASSIST Software has proven once again that it is a cradle for innovation, development, and brilliant ideas that help entire communities. New and talented software engineers also have a spot at a company that sincerely focuses on their progress and learning.  

In 2022, Best Innovative Minds will return!