Perception technologies used by mobile robots, based on cameras, LiDAR or RADAR can detect obstacles in the robot’s path during navigation. However, the challenge lies in accurately differentiating between traversable and non-traversable obstacles. LiDAR alone can detect soft vegetation, like tall grass, and identify it as a non-traversable obstacle even though it might not affect neither the vehicle's movement nor safety. This paper introduces a mechanical device called a "Sensitive Bumper Probing System," which can be integrated into mobile robots and interact with the environment to help differentiating physically traversable from non-traversable obstacles and ensure the system's safe operation. To achieve this goal, the sensitive bumper provides force measurements as it interacts with the environment. Combined with a 2D LiDAR, this measurement permits to decide whether the detected obstacles are traversable and collects data about objects to contribute to the environment recognition.