The current number of teams is 15. This is surely not the final count, because, as every year, not all the teams manage to finish their robot in time. You can look forward to see experienced „troupers” as well as „newbies”, approximately in 1:1 ratio.

Teams like ARBot, Eduro Team, Radioklub Písek, Roboauto (2x), Short Circuits Prague and Smelý Zajko you could have already seen on Robotour contest before, mostly many times. That does not necessarily mean that the robots will be exactly the same with the same brains…

Vizualizační SW

RoboOrienteering 2011

ARBot is a small robotic vehicle with dimensions 28x28x50 cm. It is based on boogie 4WD1 from Lynxmotion with 4 motors EMG30, off-road wheels with diameter 11 cm. The heart of the robot is a powerful DSP BF537, which provides 1000 MIPS computational power. There is 32 MB RAM available, 4 MB Flash, camera OV9655, 3 sonars SFR08, GPS LEA-5H, AHRS VN 100, secondary 3-axis compass and accelerometer. The control of motors is managed by two MD25 units. The communication with the external world is by WiFi, OLED touch display and piezo speaker. The energy source is 10 NiMH batteries pack with 4.5 Ah capacity. The robot is equipped with a holder for 5l beer barrel.

The control program for the robot is written in C language and some time critical routines in assebler. The road detection is computed from histogram segmentation in HSV color space of grabbed image. The control algorithm tries to keep robot in the middle of the road. The objects with different colors are not considered to be road and this automatically leads to an obstacle avoidance. Sonar is also used for the obstacle detection, its avoidance and stop in case of emergency. Readings from encoders, AHRS and GPS are used for robot position location and orientation.

The winners of Robotour 2010. There are almost no hardware changes of the robot since the last year. Learning from past mistakes, we replaced the two-dimensional magnetic compass with a three-dimensional one. Other sensors remain untouched: Odometry, lidar (SICK LMS100), sonar (not used), 1Hz GPS and an IP camera with a fish-eye lens. The ALIX single-board computer takes care of all computations. If we need more computational power, we have Gumstix Overo Fire boards ready to take over part of the work. We are getting the software where we wanted it last year. There is a big chance we break it doing this We focus on road recognition and obstacle avoidance. We may even use the GPS information this year. However, who would bother, whent it is better without it? (JI)

What's new in Radioklub Pisek? „We had to modify the robot due to the rules change - there is a new holder of beer barrel and the construction is firmer now. The robot is lighter by one piece of battery and we improved connectors for easier and faster battery exchange. The H-bridges are modified and we are working on a new motor controller. We also work on new software, so there is still what to do, and we hope we will manage everything all in time to our satisfaction.”

Karlik is modified toy car for kids PegPerego with modified accelerator controller from MAN truck. Smooth speed control forward and backward, LIDAR sensor for the obstacle detection, inertial unit, GPS, ultrasound detector in front, magnetic compass, Sony camera. The power source is 12V battery from car. Computational power is based on two notebooks Hewlett Packard + 100 MBit switch.

SICK LIDAR + ultrasound + camera are used for obstacle avoidance. Software is all home made, based on Windows platform, combination of applications in C++, Delphi and Java. The intercommunication is via TCPIP.

Quido

Quido

Quido is modified car hobby model. Implemented is smooth speed control forward and backward. There are 3x LIDAR sensors for obstacle detection, inertial unit, GPS, magnetic compass and camera.

The software is written in C++, Delphi a Java. It runs on Windows OS and it communicates via TCPIP.

The robot is based on platform from 1:5 model of monster truck. The main computational power is standard notebook. The hardware is monitored via Arduino Duemilanove board. The robot's sensor set consists of magnetic quadrature encoder, compass, sonars, GPS receiver and camera. Software is written in Python language. The program collects current information about HW status via USB. This information is gathered by microchip ATMEGA328, and regularly sent as a data packet to the main computer. The main program ensures status transformation (independently on HW platform), PID regulator of speed, localisation from odometry, GPS and compass, obstacle detection with corresponding actions and finally navigation on entered trajectory. The robot recognizes road from camera images and detects obstacles with sonars. Information from both sensors integrated into local map and with Vector Field Histogram algorithm plans next step.

At the moment we are only a step forward. We managed to get working new compass, tested various types of neural networks, improved formulas for robot vision, and it is driving more or less fine. There is still plenty of space for testing and further improvement …

There are several new teams, but their information is quite incomplete. There is typically still a lot of work on robotic platform so mostly even photo of the robot is missing now. Hopefully that will change soon.

[FAIL] is an Austrian team. The route decryption will happen on a notebook and the basic controls will be handled by a microcontroller (likely an ATMEGAXXX, 32bit are also possible though) The chassis will most likely be based on a track vehicle The Sensors will use both, optical and ultrasonic principles for obstacle detection. If there is time left, obstacle avoidance could very well be possible. Part of the Software is going to be written in C++ and maybe also Basic, the Controller Firmware in C.

Team from Poland, experienced participants from Robot Challenge.

Slovak team, with a custom built robot base with mostly custom built electronics. Robot is controlled by a netbook. The robot will use ultrasonic range finder for obstacle detection. Software is based on ROS with custom code running under Linux.

This is spin-off of Radioklub Písek team. This team will compete with four wheel drive base Dřevo from RK Písek and it will try to avoid obstacles with sonar.

Austrian team from Fh Technikum Vienna: Our roboter design is based on the Pioneer 3-AT. Apart from the 16 ultra sonic sensors which a pre build in, we have added a i5 quad core to do all the main work. Sensors wise we have added a bumblebee stereo camera, a laser, an odometrical unit and gps. The programming was done under Linux and is based on ros as well as self coded routines.

Slovak team based in Bratislava.

Team members are mostly software engieneer and theoretic robotic who finally decided to build their own robot. Because of the mostly software background the focus will be mostly on control software, architecture, extendability and robustness.

The robot name is "Spirit j.r." ( i.e. Junior), and it is 6 wheels robot based on 1:8 rock crawler, 3 cameras, Kinect, 4 ultrasound sensors, compass and GPS.

New Czech team, which built robot on Audi TT hobby model platform.

HARDWARE:

For obstacle detection we will use IR Sharp rangefinder and ultrasound sensor. The code is written in C# and in C.

The team from Slovenia FERI Maribor and ERŠ Ptuj. For basic platform we built a robotic platform in shape of the tank, which is universal and will also be used in other projects. For purpose of this competition we mounted portable computer, SICK 2D laser rangefinder, camera and GPS on the top of the robotic platform. On the computer we are running software partly written in Visual C# which is taking all the decisions and controlling the robot and partly in Matlab for video recognition. All the other electronics is mounted in robotic platform.

If you want this contest somehow support or you have some comments/questions, do not hesitate to contact us via web form.