Holonomic drive train and field-based control

[Lil' Lavery]

Originally Posted by T3_1565

how well do you think an omni bot would drive without sensors? Like a frame omni wheels and the motors (and obviously programmed). has anyone actually tried it without encoders or gyros? (our buget it limited A LOT)

It can drive quite well, you would just have to account for any mechanical inefficiencies in your code and have a "robot-centric" control system. Just like any open-loop system, your driver would have to account for anything that goes wrong and anything that doesn't work properly.

[Lil' Lavery]

Originally Posted by dmlawrence

What is the approximate cost of the materials needed?
How significant is the reduction in pushing power?
What are the (dis)advantages of field-based control?
Overall, is it worth it?

1) It depends on the exact method you select, but it can be done fairly cheaply. With some modification to kit parts it would be possible to build one for only around $250 beyond the basic KoP. Although it could range to much higher depending on the specific wheels, transmissions, sensors, etc. you select.

2) Once again depends on the method of holonomic drive you select. In the common 4-wheeled method, you get between 50-70% of your torque (varying on the angle you're driving at).
Additionally most currently available omni-wheels (and mecanum) have significantly less traction than other wheels (particularly traction wheels). But this is a problem that could potentially be solved (although the torque reduction is something inherent with the system itself).
Higher traction omni-wheels (and mecanum wheels) often have enough traction to allow for a robot to hold it's position, but rarely enough for significant force to dislodge an opposing robot from it's position.

3)The biggest advantage of the field-centric control is it allows accurate "Frisbee motion" (translation and spinning simultaneously) in straight lines to be accomplished without significant strain on the driver. It also may aid your driver in other situations as s/he doesn't have to account for the current orientation of the robot.
The high contact environment may also cause issues with the gyro, and a field-centric system makes it harder to complete "frisbee motion" in arcs (although this takes a great deal of programming to accomplish this period).

4)In the past holonomic drives have not had a great deal of success. To the best of my knowledge, only 341 (2007) has won a regional using a holonomic system (and a handful using mecanum). Although at least part of this lack of success also stems from the far lower quantity of teams that use these systems. It would be interesting (although next to impossible) for a numeric study to show the success rates of robots featuring holonomic and mecanum systems (especially considering that in the early days of these systems only teams with more resources used them).
The basic trade-off added maneuverability for torque (and often traction and ability to traverse inclines and other terrain features). That must be analyzed with each new game, but to this point it seems like holonomic hasn't granted a huge advantage in any previous games.


I'd highly recommend that you browse some of the other threads on Chief Delphi dealing with holonomic/mecanum systems to get further information.

[Billfred]

Add to your list of regional champions 343 in 2002. The pictures don't show it particularly well, but six omniwheels (two omniwheels on two sides, one on the others) drove that robot to a victory at the NASA KSC Southeast Regional alongside 180 and 186.