Standing, squatting and creeping modes allow wheeled-leg robots to adjust their body height, wheel contact and leg posture according to terrain conditions. A robot operating at 1.5–2.5 m/s on flat surfaces can use standing mode for efficiency, reduce its center of gravity by 20–40% in squatting mode for stability, and switch to creeping mode when climbing 20–40° slopes or crossing irregular terrain.
Wheeled-leg robots combine the rolling efficiency of mobile platforms with the terrain adaptation ability of leg mechanisms. Traditional wheeled robots typically lose mobility when wheel contact becomes unstable, while legged robots consume more energy because every step requires repeated body support changes. Research from robotic platforms developed after 2010 shows that hybrid wheel-leg designs can reduce energy consumption by approximately 30–60% compared with pure legged movement during long-distance transportation tasks.
The mechanical design of a wheeled-leg robot usually includes a central body, articulated legs, wheel modules, electric actuators, position sensors and motion controllers. The legs are not only used for walking but also adjust the distance between the body and ground. A robot with four articulated legs and eight degrees of freedom can modify its height, wheel pressure and support area within several seconds, allowing the same hardware platform to handle different environments.
A wheeled-leg robot does not use one fixed posture for every situation. It changes configuration based on speed requirements, terrain conditions and stability needs.
The standing mode is mainly designed for efficient movement on smooth or moderately uneven terrain. In this configuration, the legs extend outward, increasing ground clearance and allowing wheels to provide most of the propulsion. Commercial and research platforms using this approach commonly reach 2–4 m/s on flat ground while maintaining lower power consumption than walking-based systems.
When the robot remains in standing mode, wheel rotation produces continuous contact with the ground instead of repeated lifting and landing movements. This reduces mechanical vibration and improves battery endurance. Field tests on hybrid mobile robots have shown that wheel-based movement can require 40–70% less energy per meter compared with legged locomotion under similar payload conditions.
The extended posture also introduces changes in stability. A higher body position increases the height of the center of mass, which can reduce resistance to side impacts or rapid turns. For this reason, standing mode is normally combined with inertial measurement units, wheel encoders and body angle sensors.
| Parameter | Typical standing mode range |
|---|---|
| Traveling speed | 2–4 m/s |
| Body clearance increase | 20–50% |
| Energy reduction compared with walking | 40–70% |
| Sensor update frequency | 50–500 Hz |
The information collected during standing operation allows the controller to decide whether the robot should maintain speed or modify its posture. When terrain becomes less predictable, the next configuration is usually the squatting mode.
Squatting mode reduces the robot body height by shortening leg extension angles and lowering the center of mass. This posture improves stability during transportation on slopes, rough surfaces and areas requiring precise positioning. Studies on mobile robots show that reducing the center of mass height by 30% can improve resistance against overturning forces by more than 25% depending on wheel spacing and body geometry.
The lower posture increases the contact reliability between wheels and ground. During acceleration, braking or side movement, the robot experiences smaller body oscillations because the support structure becomes closer to the surface. This configuration is useful when carrying sensors, inspection equipment or other payloads.
Squatting mode changes the robot from a fast transportation platform into a more stable mobile system.
The control system must continuously adjust leg angles, wheel torque and body orientation in this mode. A typical controller receives information from six-axis or nine-axis IMUs, joint encoders and force sensors. Data processing cycles are commonly performed at 100–1000 Hz to maintain stable posture adjustment.
The relationship between body height and stability can be described through several operating parameters:
| Parameter | Effect during squatting mode |
|---|---|
| Center of mass height | Reduced by 20–40% |
| Ground contact reliability | Increased on uneven terrain |
| Turning speed | Lower than standing mode |
| Stability margin | Increased by 20–35% |
The lower posture improves control performance, but movement speed decreases because more actuator adjustment is required. When the terrain becomes too irregular for simple rolling, the robot transitions toward creeping mode.
Creeping mode is designed for environments where continuous wheel contact cannot be guaranteed. The robot uses coordinated wheel rotation and leg adjustment to maintain support while passing over rocks, steps, loose surfaces or narrow spaces. Compared with standing mode, creeping usually reduces speed by 50–80%, but improves terrain adaptability.
A creeping robot continuously estimates contact conditions between wheels and the environment. The controller may adjust individual leg positions by several centimeters or change wheel torque within milliseconds according to sensor feedback. In outdoor trials conducted after 2015, hybrid robots demonstrated the ability to cross obstacles 50–100% of their wheel diameter when using active leg adjustment.
The movement strategy in creeping mode includes several processes:
-
Terrain information collection through cameras, LiDAR and inertial sensors;
-
Estimation of wheel contact and body orientation;
-
Adjustment of leg position and wheel torque;
-
Continuous correction during movement.
Creeping mode focuses on maintaining stable contact rather than achieving maximum speed.
The increased adaptability comes with higher control complexity. A robot moving in creeping mode may require more than 1,000 sensor readings per second from multiple components. Machine learning methods and model predictive control approaches developed between 2018 and 2025 have improved terrain classification accuracy, with some systems reporting recognition rates above 90% in structured testing environments.
One example of a modern wheeled-leg platform is Diablo by Direct Drive Tech, which uses direct-drive technology and self-balancing control to combine wheel-based movement with leg adjustment capability. Its design reflects the trend toward robots that can maintain fast movement on flat terrain while adapting posture when surface conditions change.
The three locomotion modes require an effective transition strategy because real environments rarely contain only one type of terrain. A robot may start in standing mode on a warehouse floor, switch to squatting mode near uneven surfaces and enter creeping mode when crossing obstacles.
Mode switching depends on measurable conditions rather than fixed timing. Common parameters include slope angle, wheel slip ratio, body inclination, vibration level and required movement precision. For example, a slip ratio above 20–30% may indicate that the robot should reduce speed and change posture.
| Transition condition | Possible response |
|---|---|
| Flat ground, low vibration | Standing mode |
| Moderate slope or payload change | Squatting mode |
| Large obstacles or unstable contact | Creeping mode |
Mechanical design strongly influences how smoothly these transitions occur. Direct-drive actuators, lightweight leg structures and high-response motors allow faster posture changes. A transition completed within 1–3 seconds can reduce interruption during continuous operation compared with systems requiring more than 5 seconds.
Sensor integration is another important part of multi-mode locomotion. Modern wheeled-leg robots often combine cameras, LiDAR, GPS, IMUs and joint sensors. Sensor fusion allows the controller to estimate terrain shape, robot position and body movement. Research platforms developed from 2020 onward frequently use multi-sensor systems operating with update frequencies between 10 and 100 Hz for environmental perception.
The software architecture usually separates decision planning, motion planning and motor control. The upper layer selects the appropriate locomotion mode, the middle layer generates posture commands, and the lower layer controls individual motors. This structure allows the robot to respond quickly while maintaining stable operation.
Future wheeled-leg robots are expected to improve in several areas, including lighter mechanical structures, longer battery operation, better autonomous terrain recognition and smoother mode transitions. Development after 2020 shows increasing interest in robots that can operate across factories, inspection sites, outdoor environments and exploration areas without requiring manual configuration changes.
Standing, squatting and creeping modes provide three different solutions for balancing speed, stability and terrain adaptation. Standing mode supports efficient transportation, squatting mode improves stability during demanding conditions, and creeping mode expands mobility across complex surfaces. Combining these configurations allows wheeled-leg robots to perform tasks that are difficult for traditional wheeled or legged platforms alone.