Ballbot
The first successful ballbot was developed in 2005[6][7][8] by Prof. Ralph Hollis of the Robotics Institute at Carnegie Mellon University (CMU), Pittsburgh, USA and it was patented in 2010.
Prof. Hollis and his group at CMU demonstrated that the ballbot can be robust to disturbances including kicks and shoves, and can also handle collisions with furniture and walls.
[18][19][20] They also demonstrated the ballbot's capability to autonomously navigate human environments to achieve point-point and surveillance tasks.
[33] Tomás Arribas (Spain) developed the first ballbot using LEGO Mindstorms NXT in 2008 as the Master Project at University of Alcala.
Yorihisa Yamamoto (Japan) inspired by Tomás Arribas's project, developed a ballbot using LEGO Mindstorms NXT in 2009.
[40] A group of students from ITMO University (Russia) introduced an algorithm and constructed a ballbot based on LegoNXT robotics kit which performed stability with only two actuators used.
Equipped with three motors and omniwheels, an onboard Intel NUC, two SICK LiDARs, an ARM microprocessor and a tablet on the top, the robot is capable of maneuvering autonomously in indoor environments and guide people around.
The complete master thesis and all material including MATLAB source code and C++ controller implementations are publicly available on GitHub.
The wide base makes it difficult for statically-stable mobile robots to navigate cluttered human environments.
The most fundamental design parameters of a ballbot are its height, mass, its center of gravity and the maximum torque its actuators can provide.
The choice of those parameters determine the robot's moment of inertia, the maximum pitch angle and thus its dynamic and acceleration performance and agility.
Therefore, friction coefficients of all parts involved in force transmission also play a major role in system design.
[32] The ball is the core element of a ballbot, it has to transmit and bear all arising forces and withstand mechanical wear caused by rough contact surfaces.
Rider[27] used a basketball, while BallIP[29] and Adelaide Ballbot[40] used bowling balls coated with a thin layer of rubber.
In order to achieve yaw motion, the CMU Ballbot uses a bearing, slip-ring assembly and a separate motor to spin the body on top of the ball.
Prof. Masaaki Kumagai,[3] who developed BallIP[29] introduced another ball drive mechanism that uses partially sliding rollers.
[22][23] Conversely the Kugle robot[51] uses two SICK TiM571 2D LiDAR to localize itself, perform obstacle avoidance and detect people for guidance.
[24][25][26] The mathematical MIMO-model which is needed in order to simulate a ballbot and to design a sufficient controller which stabilizes the system, is very similar to an inverted pendulum on a cart.
The inclination angles of a ballbot is thus dynamically linked to the resulting accelerations of the ball and robot leading to an underactuated system.
[22][23] Its motion planner plans in the space of controllers to produce graceful navigation, and achieves point-point and surveillance tasks.
A path-following strategy is chosen over common trajectory or reference tracking controllers to accommodate for the temporally lacking behaviour of ballbots due to the underactuated nature.
The CMU Ballbot[8] introduced three retractable landing legs that allow the robot to remain standing (statically-stable) after being powered down.
[13][15] Rezero featured a roll-over safety mechanism in order to prevent serious damage in case of a system failure.
[32] Due to its dynamic stability, a ballbot can be tall and narrow, and can also be physically interactive, making it an ideal candidate for a personal mobile robot.
[8] It can act as an effective service robot at homes and offices and offer guidance to people in e.g. malls and airports.
