The castors for the front and back of the robot need work. In the current design the castors are simply 3D printed hemispheres. This made it easy to get the right height, however there is zero margin for error. Slight slopes can cause the bot to get stuck. Rugs and carpets are a no-go.
I tested both 3D printed and purchased nylon ball castors but neither offered much improvement. We could use wheel castors but then we would need to raise the robot off the ground more meaning bigger wheels. Without cutting holes in the top plate we have about 5mm extra we can go which probably isn't enough. We could possibly cut mounting holes for the wheel castors to mount them to the top rather than bottom side of the plate to gain 6-7mm.
A more complicated option would be to copy robot vacuums which have a slight suspension on the wheels allowing them to gain grip even when its slightly uneven.
A simpler option might be to redesign the 3D printed castors to have an optimised angle.
The castors for the front and back of the robot need work. In the current design the castors are simply 3D printed hemispheres. This made it easy to get the right height, however there is zero margin for error. Slight slopes can cause the bot to get stuck. Rugs and carpets are a no-go.
I tested both 3D printed and purchased nylon ball castors but neither offered much improvement. We could use wheel castors but then we would need to raise the robot off the ground more meaning bigger wheels. Without cutting holes in the top plate we have about 5mm extra we can go which probably isn't enough. We could possibly cut mounting holes for the wheel castors to mount them to the top rather than bottom side of the plate to gain 6-7mm.
A more complicated option would be to copy robot vacuums which have a slight suspension on the wheels allowing them to gain grip even when its slightly uneven.
A simpler option might be to redesign the 3D printed castors to have an optimised angle.