A drivetrain consists of all of the components used to make the robot move around the field and is a crucial part of a robot. Most drivetrains use motors to spin wheels or drive treads (belts). Some innovative teams also incoporate pneumatics and servos to allow shifting of gears or changing which wheels are in contact with the ground.
Drivetrains can come in many different forms, each one having its own strengths and weaknesses. This overview will attempt to give an general synopsis of drivetrain do’s and don’ts, as well as touch on the most common drivetrain variations.
The optimum drivetrain for a robot is one that finds the correct balance of speed, maneuverability, and pushing power to fit within a team’s strategy. Speed and pushing power are, in most cases, inversely proportional to one another and can be adjusted by modifying the gearing ratios between the input (motors) and the output (wheels).
Maneuverability can be defined as how easily it is for a robot to maneuver across a surface such as the game field. Maneuverability is much trickier to manipulate as it is the result of not only a robots speed and available torque, but of the drivetrain’s geometry as well.
A high level of maneuverability is very important in robot design, although it can be just as much a curse as it is a blessing if not correctly understood. A robot that is both extremely fast and extremely maneuverable would be almost impossible to control without a certain level of sophistication within its controls which can take the form of either a closed loop software control (e.g. PID feedback loop), or an extensive open loop control system.
Types of Drivetrains
Two Wheel Drive
The two wheel drive (aka “casterbot”) is undeniably the easiest one to make. A two-wheel drive consists of having two powered wheels at one end of the robot and a unpowered surface that slides on the other end of the robot.
The two wheel drive offers high maneuverability while maintaining a very low level of complexity. It can also be the most difficult to use. A casterbot turns with such ease because there is virtually no side friction working against the wheels of the robot as it turns. However, that same ease of turning can be a nightmare - with no friction regulating the bot’s speed of turning, its inertia will always want to continue turning even after the motors have stopped. This results in a robot that is extremely difficult to control using basic control methods - it is a rare casterbot that can drive in a straight line without the use of internal sensors or gyros. Furthermore, the unpowered surfaces in contact with the floor detract from available pushing power, and significantly lower a robot’s ability to maintain its position when hit.
The design, for all its shortcomings, is nevertheless viable if correctly put to use. When designing a two-wheel drivetrain, the powered wheels should be in the center. This allows the robot’s pivot point to remain close to its center of mass, minimizing the area through which it must travel in order to turn. Placing the powered wheels along the 30” sides of the robot will further slow its rate of turn, making it more controllable (helpful, but not necessary).
Four Wheel Drive
The four wheel drive system is probably the most common drivetrain used in FIRST. It offers a number of advantages and disadvantages over the more basic two wheel drive. With the addition of two extra driven wheels, a four wheel drive robot has more traction and control over a two wheel drive robot. The trade-off is the increased wheel base can cause problems turning (when the wheel base exceeds the wheel width).
Six Wheel Drive
The six wheel drive is a moderately common type of drivetrain. It offers a good compromise between traction and maneuverability. Most teams choose to lower the center wheels by approximately 1/8 - 3/16”, allowing the robot to turn more easily because normally it has a shortened wheel base but at the same time, can tip over slightly, and enjoy the benefits of a full length wheel base.
Some six wheel drive robots are in fact four wheel drive robots with an unpowered pair of extra wheels that just act to increase the wheel base for stability. When tipped to the unpowered set, some of its weight is diverted to unpowered wheels and traction is therefore reduced.
A variety of a Six Wheel Drive is the West Coast Drive, unsurprisingly commmonly found on the west coast with teams such as the Cheesy Poofs and features live axles, open sides and chain linked wheels.
A car drive (aka Ackerman Steering) robot has a steering system much like what would be found on a standard automotive vehicle with front wheel steering.
Usually, the drivetrain system has four wheels, with the two wheels in the back providing power, and the two wheels in the front providing steering, though there are designs with power and steering to all wheels. While the design gives increased speed and pushing power and reduces the learning curve for the driver, it is not a very common choice due to its lack of maneuverability. Generally, the large turning radius makes it very difficult to maneuver in a corner or tight space.
A swerve drive robot has the ability to rotate its wheels, allowing the robot to travel with three degrees of freedom.
Several types exist, such as the crab drive, in which all four wheels are linked such that they always have the same angle as each other (generally requires one drive motor per wheel plus two angling motors). A 2+2 configuration has two pairs of wheels which share angles (typically requiring four drive motors plus two angle motors), and a full omni system allows each wheel to be independently angled (requires four drive motors and four angle motors.
The main advantage of a swerve drive is a great increase in maneuverability. The trade-offs are that swerve drives are much more complex to build and consume much more resources (time, money, weight, space) than most other drivetrains. Some forms of swerve drive are also known to have less power for pushing other robots around on the field. However, increased maneuverability is gained to make up for it.
A crab drive allows the robot to strafe at any angle, but turning involves a great deal of skidding, sometimes even more than a typical four-wheel tank drive system (in which the power to the right and left wheels can be independently adjusted). A 2+2 configuration allows smoother turning (similar to a car drive system but with two sets of independently angled wheels) and the same omnidirectional strafing as a crab drive, but it cannot turn and strafe at the same time (the reasons for this are a bit complicated). Finally, a full omni system allows for any combination of strafing and turning simultaneously, as well as point turning. However, the increasing weight and complexity of the latter two designs leads many teams to use the simpler crab drive.
A holonomic drive (aka “omni drive”) also allows a robot to travel with three degrees of freedom rather than two. The difference is that a holonomic drive allows a robot to instantly change direction without having to turn the wheels to a different position. The major benefit of using a holonomic drive is a great increase in maneuverability without having to add an entirely new mechanism to turn the wheels. The major trade-off is that the robot isn’t very good when it comes to a pushing and shoving. Holonomic wheels have poor traction, as they can’t be made inflatable or with treads. They also demand individual, speed controlled motors for each wheel.
Here is a video of team 1418’s 2007 holonomic drive in action. Note the slight listing when the chassis is supposed to be driving straight; this was caused by asymmetries in the power outputs of the Victor speed controllers. This problem was successfully fixed by using a lookup table to force individual joystick positions to map to a linear set of power outputs.
A mecanum drive is another omnidirectional drive system. It consist of wheels with their rollers angled in a conventional four wheel drive layout. With an independent motor/transmission on each wheel, omni driving can be achieved by varying speeds.