Re: [beam] Plans for the future...
- And with a slight slip of the finger, i make a fool out of myself :)Please disregard this email.David----- Original Message -----From: David PerrySent: Saturday, 22 July 2000 1:46Subject: [beam] Plans for the future...Hey all,Well it seems this years attempt to enter robocup has crashed and burned, probably because Bruce is always right. :) Time to regroup.It seems to me that we had an organsational problem. As Andrew H. pointed out, we have excellent resources and skills to work with - optics, electronics, embedded controllers, programming, mechanical engineering, etc. The problem is integration and organisation. As i recall we do have 10 members, and only 4 have contributed anything. The answer (IMHO) goes like this...There shall be be 3 systems groups, Mechanical, control, and vision.* Mechanical systems group is concerned with the mechanical aspects of the robots which includes motors, wheels, base, kickers, motor drivers and some sensors such as encoders.* Control systems group works on the control systems, the central portion if you like. This includes behaviour design and implementation. Also includes some sensors such as bump sensors. Also includes communication.* Vision systems group works on vision (duh), which includes 2 sub groups, BEAM vision and conventional vision. The idea being that if BEAM based vision does not work or doesn't work well, the conventional camera based system takes over. And if the BEAM system is effective, the conventional system may be used as well for experimentation.The conventional system BTW will be based on a color digital camera (Melb. Uni. is designing an open source camera which should be suitable) connected to some kind of CPU, probably a few of them in fact to handle the amount of processing required.Each group has members which work on that system, the group itself designs the system, and each member works independently with one aspect of that design. Each member reports to their own group every week. And any problems are discussed amoung that group.The exception of course is any thing that meshes with another system, in which case, the beam_soccer group is used for communication. Each systems group also sends an update to beam_soccer every couple of weeks.Sponsorship is also an issue.We have a sponsor, but in reality we also need funds for prototyping as well, and the benefit for RobotOz fades with the competition being in Seattle. A PR campaign will help with this, we need to promote our team in order to find another sponsor, a proper public website is in order.Weekly organised chats that as many people as possible attend are also necessary.Another recruiting drive may be necessary as most of our team mates don't seem to do anything...What are your thoughts?David