Projects

Brief information on the projects I have taken part in is given below. To reach my coursework projects click here.


Integrating Combinatorial Reasoning and Continuous Methods for Optimal Motion Planning of Autonomous Vehicles

As of December 2013, I have been working on my master's thesis at the department on Mobile Perception Systems of FZI Research Center for Information Technology under supervision of Ph.D. cand. Philipp Bender. My thesis is about optimal kinodynamic trajectory planning for autonomous vehicles. To gain insight into my research you can watch the video below and refer to my master's thesis.

For more information, you can also refer to the papers published by our group: Trajectory planning for Bertha - A local, continuous method , Lanelets: Efficient map representation for autonomous driving .



Porsche InnoDrive: Online estimation of driving resistance parameters

As a part of the next generation Porsche InnoDrive technology, collaborative research is done between Dr. Ing. h.c. F. Porsche AG and KIT - Institute of Vehicle System Technology with the goal of estimating the unknown driving resistance parameters by the use of online filters.

My advisor in this project Ph.D. cand. Stephan Rhode is researching online filter types and is implementing them in MATLAB and finally performs test-runs to investigate the performance of the vehicle models obtained, with the use of acquired data from the vehicle CAN-Bus. Thereby, more accurate range prediction and driving strategy can be obtained.

My duty in this project was to provide assistance to my advisor in post-processing. Especially creating codes in MATLAB for various filter types, plotting figures which could serve for the ease of model selection, improving the visualization of figures, and literature review about the topic.

To get more information about the Porsche InnoDrive project, you can watch the video introducing the project below:



MEKAR Grand Cooperative Driving Challenge Vehicle's Automatic Brake & Gas System Design

The Grand Cooperative Driving Challenge (GCDC) is an international challenge that aims to accelerate the development of cooperative driving technologies. Cooperative driving aims to form platoons of vehicles on the roadways by having the cars communicate with each other and with the infrastructure. By transmitting the vehicle's acceleration information to the vehicles behind it, the vehicles inside a platoon will be able to brake and accelerate simultaneously. The eliminated human reaction times will allow a closer headway between vehicles and thereby reducing it significantly. Hence, with the use of Cooperative Adaptive Cruise Control (CACC), a more efficient use of existing infrastructure alleviating traffic problems and reduced fuel consumption will be achieved.

The GCDC is held by TNO in the Netherlands on the 14th and 15th of May. Team MEKAR has challenged its rivals with its half-autonomous car and placed 7 th overall and 5thin the individual category. It was a good rank for us as we did not have any autonomous or semi-autonomous vehicles which we could easily equip with the neccessary hardware and software required for the C-ACC. The vehicle we had developed was not fitted with an OEM ACC, so we also had to deal with incorporating Adaptive Cruise Control (ACC) into the vehicle, which most of the other teams had as stock on their vehicles. Our vehicle was intrinsically very sluggish due to its heavy weight, exceeding 1700kg with the equipment installed, compared with the IC engine of power of 90 hp (67kW). Besides the low hp/1000kg ratio, the delays during the shifting were also relatively high due to the tiptronic shift, which made our vehicle even slower. More information about our MEKAR GCDC Vehicle can be found on the IEEE Transactions on ITS GCDC Special Issue.

Team Mekar

Team Mekar's GCDC vehicle is a Fiat Linea sedan provided by our major automotive OEM sponsor Tofaş-Fiat for use during the race. Istanbul Okan University, İstanbul Technical University and OTAM (Automotive Technologies Research and Development Company) were our main sponsors for the GCDC event. I was responsible for designing a closed loop control system needed for braking and accelerating purposes by forming the necessary longitudinal vehicle model, and the mechanical design needed for sensor, actuator etc. mountings. Mechanical design had a constraint that we had to make the changes on the car without drilling any holes or making any weldments. I was the only undergraduate student in this project, and my senior design project covered controller integration by using a rapid controller prototyper (we were using Microautobox) and performing road tests. More information about our car can found at MEKAR GCDC Team's web page.



'Bee-R' ITU Formula SAE Car's Mechanical Design

ITU Formula SAE Team, which is Turkey's first Formula-SAE Team, was founded in 2006. Our target was to take the iniative to start Formula SAE in Turkey and produce a car for the 2007 Formula SAE Italy event. However, due to the difficulties of making the university's and also the country’s first Formula SAE car, the efforts failed to design a producible one. In October of the year 2007, remaining members of the team, mostly consisting of second year students and two M.Sc. students, started to work on a brand new design.

Due to our very small team and bureaucratic procedures, our work slowed. In the Spring of 2010 we, as ten students from different engineering branches, completed Turkey’s first and only Formula SAE Car. After four years of lasting faith and commitment to complete the project, we entered the Formula SAE Michigan 2010, event.

In this project, as we were only a few members, I (and also the other members) worked on almost the entire mechanical design. However, the chassis and suspension design were in my responsibility.

After attending Formula SAE 2010 - Michigan event, as the designers of the first car, we founded a completely new team for the design and production of a new F-SAE car for the future events. We also left the team leadership to a freshman student in order to establish a new team, but providing them with experience and knowledge. Although, we as the founding members are not in the team anymore, we all are trying to mentor the teams in fields where guidance is needed.

ITU Formula SAE Team is now designing its second car to compete in the Formula Student Germany. Formula SAE (or Formula Student in Europe, Formula ATA in Italy) is a student design competition organized by SAE. ITU Formula SAE Team is determined to produce a powerful car for the Formula-ATA, 2013 Event. More information about the ITU Formula SAE Team can be gathered from the team's web page.



'HaydARI' Minisumo Robot Design

A sumo robot's task is to push the robot that it is competeting with to outside of a circle. Mini-sumo robot is a category of the sumo robots. They can be up to 500 grams and 10 cm by 10 cm without any height limit. Sumo robot competitions are held worldwide with Japan being the pioneer. I had prepared a mini-sumo robot for METU Robotic Days in 2007. Because it was my first robot with which I took part in a competition, it has a special place for me.

'haydARI' Minisumo Robot

To determine the position of the opponent robot, five Sharp 2D150A F IR sensors were used, with there of them looking forward. The neccessary algorithms were written on ANSI C and were implemented on a PIC16F628 microcontroller. To overcome the 10 cm x 10 cm restriction, the robot was designed to expand after the start by opening its front plate.