2021

Project Name : Steer-By-Wire

Project Team : Can Acar, Aghil Ajith, Taylan Aydın, Hakan Başargan

Project Summary : Traditional mechanical driving is the biggest driving challenge. The mechanical links between the steering wheel and the wheels take up unnecessary space, cause delays when driving and transfer all road vibration to the driver, which ultimately translates into weaker driving safety and uncomfortable driving. With Steer-By-Wire, all the unnecessary mechanical links between the steering wheel and the wheels are removed, sensors and servos are added to the steering wheel and steering gear, and the cars are driven solely with digital information, much like planes. This opens the doors to an excellence in driving experience: Now the steering wheel can be in any form and in any place within the car; be it right, left, middle or as any kind of a joystick. Or it can be completely removed. As it saves place in the cockpit, the driver can be the passenger in any time. Driver seat can turn 360 degrees. With these, Steer-by-Wire makes 5th grade autonomous driving a reality. We perfect the real driving experience with 3 different control system loops, the first one of which provides steering sensitivity, the second to control the servos to drive the wheels, the third to check on the angle ratio between the steering angle and the wheel angle. Not even a millisecond of delay is experienced during the ride, and no vibration from the road is transferred to the driver. The system also reduces the weight of the car, which translates into better battery and fuel performance and lower emissions. Steer-By-Wire has a control system that continuously perfects the driving experience in view of environmental factors. By evaluating the information downloaded from the network such as road condition, speed, weather conditions and optimizing the output value on the wheels, it provides a perfect driving experience unprecedented in traditionally built automobiles. In an event of danger, Steer-By-Wire executes the necessary manoeuvres in milliseconds before the driver and can stop in shorter braking distances and shorter times no matter the condition of the road. Our competitive edge comes from removed mechanical parts and global road worthiness of Steer-By-Wire. Besides the automative industry, this system can easily be integrated into defense industry and agricultural vehicles.

The Amount of Prize: 140.000 TRY

Project Name : HydroBorPEM Fuel Cell Technologies

Project Team : Erhan Demircioğlu

Project Summary : The ever-increasing pollution and the high possibility of running out of fossil fuels drive the tendency to switch to automobiles powered by alternative energy sources. Presently, as one of the automobiles powered by alternative energy sources, vehicles with hydrogen fuel cells come to the fore with their efficient and environmentally compatible technologies. HydroBorPEM manufactures 6 times efficient Boron and Perlite doped composite fuel cell membranes and fuel cell stacks for hydrogen fuel cell vehicles.

 

The Amount of Prize: 120.000 TRY

Project Name : Advoard Robotics

Project Team : Pınar Oğuz Ekim, Sercan Çağdaş, Tekkök Mehmet Emre Söyünmez, Bekir Bostancı

Project Summary : In logistics and e-commerce warehouses, the multi-item shipments are carried out by workers strolling through the warehouse, collecting the item and bringing it to the cargo unit. Which is a physically tiring task that requires attention. Losing focus and insufficient training can lower the efficiency of the collecting task. The workers spend 90% of their time looking for the item in the warehouse and 10% of their time actually collecting the item. The warehouse structure cannot keep up with the influx of shipping frequency on special occasions and days, for which temporary workers are hired, who are then trained for long periods of time. Additionally, mistakes during the collection of an item cause wrong items to be shipped. All of these problems add up to losses of efficiency, money and time. With the autonomous mobile robot we developed, we aim to solve these problems by ensuring mistake-free and efficient collecting and increase the efficiency of warehouse operations by two - three fold. Multiple robots working harmoniously with warehouse management systems and with each other will ensure the collection of items with maximum efficiency and minimum distance needing to be covered. Each one of our robots have an innovative aspect, as they map their surroundings, locate their position, plan the route and autonomously drive themselves. With the user-friendly web interface we developed, it will be possible to observe the robots, to troubleshoot, to conduct detailed fleet and collection analysis from a distance at all times. This project will make a difference as the panels placed on top of the robots will provide an easy-to-use system to the worker. More load will be carried more quickly. The autonomous mobile robots designed for e-commerce and warehouse logistics are not manufactured in Turkey. Only 6 companies work on this subject. According to their data, these companies are growing fast and their turnover is increasing. Manufactured in our country, this product constitutes an original value as it is cheaper, carries more load, is more flexible on integrations with any warehouse management system than its alternatives. With mistakes decreased and efficiency increased, the companies which use this product will obtain remarkable gains in terms of time and money.

 

The Amount of Prize: 100.000 TRY

Project Name : Hybrid Location Control Systems

Project Team : Çağatay Karademir, Mert Ali Özel, Emre Demir 

Project Summary : This project is based on combining the hardware units, cameras and software systems onboard the stations of automobile factories in order to ensure the precise determination of the location of an object. The system aims to increase the efficiency and the quality of the work and the skilled labour by preventing human errors and most importantly, time losses.

 

The Amount of Prize: 80.000 TRY

Project Name : Smart Automobile Assistant with Hologram Imagery and Artificial Intelligence

Project Team : Serdar Sultanoğlu, Erbakan Malkoç, Majed Bawarshi, Mohammed Fares

Project Summary : Our project, "On-board Holographic AI Voice Control and Vehicle Assistant System for Domestic Automobiles and All Vehicles" supported by TÜBITAK is pending R&D completion. Application for KOSGEB Cooperation support has been submitted and the our project has been deemed eligible. With project development operations ongoing, The plan is to start operations on domestic cars and other vehicles once the final product has been achieved. A working prototype has been made. Firms selling VIP cars abroad such as Mansory, Klassen are contacted. The prototype was demonstrated and a presentation was made. It was ensured that the product would be bought once it is mass produced.

AI Assistant Features

-             Facial recognition with camera
-             Emotional recognition
-             Suggestions based on emotion
-             News reading
-             Weather information
-             Sending e-mails
-             Setting reminders
-             Creating notes
-             Timed reading of the notes
-             Alarm setting
-             Voice controlling the equipment on board
-             Answering questions

Alongside such features, the AI Assistant will also conduct an image processing via the camera, recognize the person and the emotion. Upon recognizing emotions such as happiness, unhappiness, sadness, anger etc., the AI Assistant will make recommendations, saying for example, "Welcome Mr./Mrs. ..., 'You look unhappy today; would you like me to make you a cup of coffee?" It will also be able to read the daily news or weather forecast. Before speaking with the car, a wake-up word will be uttered so as to not cause a mix between speeches taking place in the car and commands given to the car. For example, when the words "Hey Darvis, I want you to send an e-mail" are uttered, the Assistant will respond "To whom?" and an e-mail will be sent upon providing the information. In order to send an e-mail, the sender will need to be clearly visible by the camera and an authentication needs to take place; otherwise the Assistant responds "I need to see you to send the e-mail." Reminders or notes can be set to be read by the Assistant on a specified time.

Assistant will be able to answer the questions with AI. A "Question and Answer" model will be followed for this. When a question is asked, a research will be conducted on a search engine and Wikipedia. Paragraphs of information that were found will be simplified employing the levenshtein model, and will be sent to Google's BERT model, to which we have added new layers, in order to find the logical answer. The answer will be read by the Assistant.

The Amount of Prize: 60.000 TRY