[KrazyCarl92]

Another way that weight distribution issues can be mitigated in a tank drive (at least front to back) is by chaining, belting, or gearing the rotation of the wheels together. It appears that this is not the case in your setup. Coupling together the rotation of the wheels by some method that can transmit the torque makes it so that in order to lose traction on one side of your drive train, the total output from the motors/gearbox(es) on that side has to overcome the total tractive force of all the wheels on that side. This is instead of it being on a wheel by wheel basis.

This may help because once you have 1 wheel slipping, it kind of screws up the rest of the experimental set up. Let's suppose the weight of the robot is 100 lbs, that your wheels have a static CoF of 1 and a dynamic CoF of 0.5, and that your wheels are not chained together. If your front wheels are supporting 20 lbs each and your back wheels 30 lbs each, then your front wheels will begin to slip early on and go from 20 lbs of tractive force down to 10 lbs of tractive force. Then since those are already slipping, your back wheels will both have 30 lbs of tractive force just before they start to slip, but that will only give you a total of 80 lbs of tractive force. If you were to chain your wheels together in this example, you would get 100 lbs of tractive force.

This wouldn't solve weight distribution issues side to side, but it is a very easy and simple way to solve the problem front to back.