Adaptability is a fundamental yet challenging requirement for mobile robot locomotion. This article presents$\alpha$-WaLTR, a new adaptive wheel-and-leg transformable robot for versatile multiterrain mobility. The robot has four passively transformable wheels, where each wheel consists of a central gear and multiple leg segments with embedded spring suspension for vibration reduction. These wheels enable the robot to traverse various terrains, obstacles, and stairs while retaining the simplicity in primary control and operation principles of conventional wheeled robots. The chassis dimensions and the center of gravity location were determined by a multiobjective design optimization process aimed at minimizing the weight and maximizing the robot's pitch angle for obstacle climbing. Unity-based simulations guided the selection of the design variables associated with the transformable wheels. Following the design process,$\alpha$-WaLTR with an embedded sensing and control system was developed. Experiments showed that the spring suspension on the wheels effectively reduced the vibrations when operated in the legged mode and verified that the robot's versatile locomotion capabilities were highly consistent with the simulations. The system-level integration with an embedded control system was demonstrated via autonomous stair detection, navigation, and climbing capabilities.

Robots capable of traversing flights of stairs play an important role in both indoor and outdoor applications. The capability of accurately identifying a staircase is one of the vital technical functions in these robots. This paper presents a vision-based ascending stair detection algorithm using RGB-Depth (RGB-D) data based on an interpretable model. The method follows the four steps: 1) pre-processing of RGB images for line extraction by applying the dilatation and Canny filters followed by the probabilistic Hough line transform, 2) defining the regions of interests (ROIs) via K- mean clustering, 3) training the initial model based on a support vector machine (SVM) using three extracted features (i.e., gradient, continuity factor, and deviation cost), and 4) building an interpretable model for stair classification by determining the decision boundary conditions. The developed method was evaluated for its performance using our dataset, and the results showed 85% sensitivity and 94% specificity. When the same model was tested on a different test set, the sensitivity and specificity slightly decreased to 80% and 90%, respectively. By shifting the boundary conditions using only a small subset of the new dataset without rebuilding the model, performance was improved to 90% sensitivity and 96% specificity. The presented method is also compared with existing SVM- and neural- network- based methods.