|
|
|
|
Eurobot
2007
|
|
|
2007-06-09
|
France
/ La Ferte Bernard
As a team of two we decided to compete with our robot MULLRATS
made for this event
The task in 2007 is to clean up an area of about
2m*3m from PET bottles, ALU cans and D-size batteries
90 seconds time and
played simultaneous against another robot
(see eurobot.org for more
information)
The
robot has been made without lathe and millingmachine.
We only used an electrical
multitool and a drill without stand
Mullrats can contain up to 10 objects
(today we only use up to 7 object-places)
and uses 14 PCB's, 9 motors and
12! Atmel microcontrollers
status of our bot in february/march
2007:
Austrobot 2007 in Wels 14.4.2007:
Mullrats contains 14 PCB's:the robot is seperated in a lot of pieces for
relieving the brain-PCB and for easier developing
|
1 |
Brain-PCB |
AT MEGA 128 @16 Mhz |
|
|
|
contains the brain of the robot, and is the
middle of the star-shaped assembly
|
|
|
|
additional contains the object detection via
IR-sensors and the optional remote control |
|
|
|
|
|
1 |
Position-PCB: |
AT MEGA 128 @16 Mhz |
|
|
|
continuously calculates the absolute position
of the robot |
|
|
|
|
|
1 |
CAM-PCB: |
AT MEGA 8 @17,2 Mhz, AT TINY 12 @1,2
Mhz
|
|
|
|
uses a OV6620 cam module for digitazing the
surroundings |
|
|
|
|
|
6 |
MOTOR-PCB: |
AT TINY 26 @16 Mhz (3 without TINY) |
|
|
|
each one uses a VNH2SP30-E chip mounted
piggyback, |
|
|
|
this chip can handle a maximum of 30A and 41V
in the right conditions |
|
|
|
|
|
1 |
Control-PCB: |
AT MEGA 16 @4Mhz |
|
|
|
controls the motors of the
gate |
|
|
|
|
|
2 |
MOTOR-PCB: |
AT TINY 2313 @20 Mhz |
|
|
|
not developed by us, it's a
copy |
|
|
|
|
|
1 |
I/O -PCB: |
ATTINY 26 @16 Mhz
|
|
|
|
simplest PCB, 4 pins are connencted to
LL-Fet's which are for pulsing four |
|
|
|
IR-Led's with higher current than the tiny
could deliver, four IR-transistors are |
|
|
|
connected to the ADC
channels
|
|
|
|
LINK to
entry |
|
|
|
|
|
1 |
Colour .PCB: |
ATTINY 26 @1Mhz |
|
|
|
uses a very cheap colour sensor for object recognition
(sorting) in the robot |
overview:
the bot is mostly empty on the bottom
and runs over the garbage objects
IR-sensors are checking the waste, and in
the right moment, a gate shuts the open front of the bot and aligns a single
object
a barrel seperated in four chapters turns the first object one
position up
there are three interlocked barrels which are able to handle the
objects from one to another -> place for 10 objects
at the rear of the
third barrel is a door which opens on demand for passing out the objects in the
recycle bins
Eurobot france:
due to technical problems we were not able to
pass the technical homologation:
back at home before eurobot we were able to
collect two points on an average of 80%
in france we were only able to make
points on the testtables but never on the match-tables :-(
getting
more than one object simultaneus in our robot was the main
problem
others:
table and objects were printed and not painted, they moved
very different
the centre of gravity of the foam-filled cans wasn't always in
the middle, they weren't always rolling down on a slope,
we didn't filled
them on the testtable back home...
resume:
- we were able to implement all of our
functions, !! WE LEARNED A LOT !!
- our concept wasn't good, the robot wasn't
able to pick up more objects at a time
- the robot was to slow
- because of building the robot only with a
multitool and drill a few mechanical parts were not solid enough
- the project was much too hard for a first
robot
- the atmosphere at eurobot was really
impressing
impressions:
typical
pit
french
finals
|
|
