Legged robots are special robots that can move with the help of “legs”. This is already possible with only one “leg” (hopping). As a rule, however, four-legged, six-legged or even bipedal machines are used. There are machines that have even more legs, but these can be traced back to the six-legged version.
An early object of investigation was the different types of horse gait. The photographer Eadweard Muybridge recorded the gallop of a horse with a specially developed recording technique and was thus able to prove photographically that in this gait, the horse touches the ground at different times with only one hoof and at other times not at all. Although the horse is in an unstable state at any given moment, it is still stable overall.
In order for walking robots to be able to perform their task, the sequence of movements and the reaction to disturbances must be precisely defined. In the case of wheels, these are axial rotations and occasional steering movements, while a machine with “legs” moves several components in a coordinated manner in space and time.
---
In machines that have at least six legs (hexapods, millipedes, arachnids with at least eight legs), there are always at least three legs on the ground, so that the machine is stable at all times. Even with quadrupeds, a reasonably stable stance can only be achieved if three legs are on the ground at a time. However, it is not possible to achieve a high running speed with such odd designs.
Pros and Cons of Legged Robots
Based on observations, it’s evident that legged robots offer superior mobility in natural terrains compared to wheeled robots. This superiority is gained from their ability to utilize discrete footholds for each foot, contrasting with wheeled robots that require continuous support surfaces. Consequently, legged robots can navigate irregular terrains by adjusting their leg configuration to adapt to surface irregularities. Moreover, their feet can establish contact with the ground at selected points, aligning with terrain conditions. Therefore, legs inherently serve as suitable systems for locomotion in uneven terrain.
In soft surfaces like sandy soil, utilizing discrete footholds can also enhance energy consumption efficiency. Unlike wheeled or tracked robots, legged robots tend to deform the terrain lesser, resulting in lower energy requirements to maneuver out of depressions. Additionally, by controlling the contact area between the foot and the ground, legged robots can minimize ground support pressure. Furthermore, the incorporation of multiple degrees of freedom (DOF) in leg joints allows legged vehicles to change their heading without slippage. Adjusting body height introduces a damping effect, decoupling terrain irregularities from the robotic body and its payload. Legged robots also have the capability to hug the terrain they traverse, enhancing balance, especially when moving across external surfaces such as pipes.
Another advantage under recent investigation is failure tolerance during static stable locomotion. Unlike wheeled robots, which experience severe mobility loss with wheel failure, legged vehicles can maintain balance and continue locomotion even with damaged legs due to their redundant leg configuration. Additionally, legs can be utilized not only for locomotion but also while the robot is immobilized. For instance, actively actuating the body while keeping the feet fixed to the ground provides an active support base for assisting manipulator motion or tool operation mounted on the body.
Furthermore, multilegged robots can utilize legs for manipulation tasks without the need for additional manipulator assemblies. For example, biped robots inspired by dinosaur structures use their tails for balance maintenance and manipulation tasks. Quadruped robots have been proposed for tasks like landmine detection and removal, where one leg serves as a manipulator arm equipped with various end effectors. These solutions contribute to system weight reduction and increased energetic autonomy, as arms dedicated solely to manipulation tasks are not required.
However, despite their advantages, legged robots still face significant limitations in their current state of development. These limitations include low speeds, complexity in construction, and the requirement for complex control algorithms. Additionally, the mechanisms are heavy due to the large number of actuators needed to move multiple DOF legs, contributing to high energy consumption.
Walking and Running Behavior
Static walking robots
Static walking occurs when a robot’s center of gravity is above the surface between the feet touching the ground at all times, so that it cannot fall over without the action of an external force. The classic walking robot consists of actuators, sensors and a computer control system. The “legs” are usually moved by servo motors in such a way that a predetermined movement program is reeled off.
The ASIMO robot moves at a maximum speed of 6 km/h, with a height of 1.30 m and a weight of 52 kg, and it requires a lot of electrical energy to do so. As a special ability, it can climb stairs.
Six-legged constructions are an ideal basis for structurally stable walking robots. This makes them suitable for movement on uneven terrain. In the tripod walk, there are three legs on the ground at all times. In the tetrapod gait, there are always four legs on the ground. In the case of walking machines with six orthogonal legs, a distinction is made not only according to the order of the leg movements but also according to the basic type of movement of the legs.
When walking on uneven terrain, it is crucial that the robot finds a safe touchdown point within its stride width (potential foothold selection area) without having to deviate too far from its main direction of travel.
Dynamic walking robots
Walking robots, which can move without an energy source, are based on a toy invented 150 years ago. It only had to be bumped and could then walk down a small slope on its own. To do this, the toy swings from right to left and swings the leg that has just been lifted forward a bit. Then it swings from left to right and the other leg swings forward.
With this design, the toy can move in an energy-efficient way and serve as a starting model for technically more advanced walking robots. In the 1980s, Tad McGeer used the principle of the pendulum implemented in this toy to stabilize movements. If the construction of the simple toy is supplemented with a “hip” or “movable feet”, then such walking robots only need energy to accelerate the moving masses and no longer to decelerate, as was the case with earlier walking robots.
