robotsystems.net
home   14 December 2005

Robotics Challenge

A workshop organised by Geoff Pegmanon of R U RobotsLimited, held on 1st December 05 at Salford University, was centred on future robotics research and development opportunities and aimed at both industrial and academic participation.

Following on from the workshop I am proposing a Robotics Grand Challenge.
David Buckley
2 December 2005

Proposed Challenge for Biped Robots
-----------------------------------

Challenge-1, as below, for humanoid robots to carry out a simple task.

Challenges-2,3,4,5,6,.... in later years - to include
                          steps up and down in the course,
                          doors to be opened and closed in the course,
                          random obstacles in the course,
                          etc

i.   This challenge doesn't rely on enabling third parties 
     such as TV companies.

ii.  Detractors may say the challenge is too easy - 
     then they shouldn't have any trouble winning should they.

iii. If the challenge was in another part of the world then the rules would
     be made harder since platforms are already available.

iv.  The fact that the USA put men on the Moon doesn't imply that the UK 
     could do the same even given the same resources.

v.   Any challenge arising from within needs to have a lot of successful
     entries if it is going to influence those without.

vi.  As set out below, Challenge-1 can be as hard as you like - whisper in 
     the robots ear "go get the can and bring it back" then sit back.
     Show off what you can do!

vii. The challenge is land based because not everyone has easy access to 
     suitable ponds, pools, tanks for water based challenges. Not everyone has
     easy access to suitable open spaces for air based challenges and no one in
     this country has access to off-planet for 'space' based challenges. (At the
     moment the market for planetary rovers is insignificant and in any case 
     rovers are land based.)

viii.Why a Humanoid Biped robot? For a robot which coexists with humans and 
     carries out tasks for and with humans then we already know a humanoid shape
     is, at least for the present, the best.

ix.  The design of a humanoid robot poses significant challenges, seemingly 
     overlooked or considered trivial by specialists in software and 
     electronics. No amount of effort on algorithms can redress basic
     deficiances of mechanical design. Whereas good mechanics can ease the 
     control problems considerably.

x.   The potential market for Humanoid robots is far larger than any other type.


David Buckley
2 December 2005

--------------------------------------------------------------------------------

Challenge-1
-----------
For Humanoid Biped robots.
The robots to walk down a short, non straight, course laid out by the 
organisers on a flat floor, collect an empty aluminium drinks can from off 
a table and return to the start and put the aluminium drinks can on another
similar table. 
Only one robot in operation at a time.
The winners being the quickest in each class and overall.
Any form of control allowed.
No umbilicals allowed.

Four classes
------------

1. Robot height up to 350mm,
   Course 4 metres long by 250mm wide,
   Table 125mm high x 150mm square.

2. Robot height 350m to 700m,
   Course 8 metres long by 500mm wide,
   Table 250mm high x 150mm square.
   

3. Robot height 700m to 1.5m,
   Course 16 metres long by 1 metre wide,
   Table 500mm high x 150mm square.
   

4. Robot height 1.5m to 2m,
   Course 32 metres long by 2 metre wide,
   Table 1 metre high x 150mm square.
   


Rules
-----
1. The course may start and end at the same table.

2. The total deviation of the course to be at least 90 degrees
     e.g. - L shaped with 90 bend;
          45 deg left + 45 deg right;
          45 deg left + 45 deg left;
          30 deg right + 60 deg left;
          etc.

3. The course to be marked by 2inch wide black tape.

4. The time of the run to end when the can is stood upright on the table
   at the finish and the carrying effector is clear of the table.

5. If the can falls off a table the robot may pick it up,
   human intervention is not allowed.

6. Internal combustion engines not allowed.

7. Height to exclude aerials.

8. Competitors may choose to enter a higher numbered class at their discretion.
   (I can't think of why anyone would want to do this)
 
9. Robots deemed dangerous by the judges may be prevented from competing.

10. Official tables will be rectangular prisms painted white.

11. Competitors may supply their own tables marked as they wish.

12. Tables may not contain active markers such as InfraRed emitters, etc.

13. A robot knocking over a table will be judged to have come last.
    Unless it can pick up and replace the table in it's original position and 
    then carry on.

14. Maximum time allowed will be 5 minutes.

15. Control systems can be anything from simple direct radio control, 
    (27MHz or other land approved channels),
    with a human operator,
    to anything the handlers can dream up.

16. Competitors will be allowed space to one side of the course,
    in a pits area determined by the judges,
    for any computers, chargers and minor repairs required.
    One 13A socket will be provided.

17. Mains-electricity based computers will not be allowed in the start area.

18. One minute will be allowed for any control equipment to be set up in the
    start area, after which the run will be deemed to be aborted.

19. Only two such aborts will be allowed.

20. All rules to be at the discretion of the judges in the interest of having
    an enjoyable challenge.

21. The object of the challenge is to win and have some fun.


Revisions
---------
14-12-05 'Notes' heading changed to 'Rules',
         Rule 13 revised to allow picking up of a knocked over table.


Comments
--------
1. Robots can be built to class-one for under 100 material cost. 
   (can not must)

2. Robots can be built to class-four for under 1000 material cost. 
   (can not must)

3. If anybody thinks the challenge is too simple then it should be child's play
   for them to knock something up over a weekend and show the rest how it should
   be done.

4. If everyone at the workshop entered a biped, (50 bipeds?) it would make an 
   impressive showing and, providing the engineering looked respectable rather 
   than the usual 'Bits Stuck Everywhere' approach, could serve usefully in 
   influencing the public. Also afterwards the robots could then serve as 
   platforms/tools for further research.

David Buckley
2 December 2005

--------------------------------------------------------------------------------