Department Publications

Vision-Language Navigation (VLN) enables robots to navigate through environments using natural language instructions, making human-robot interaction intuitive. Traditional VLN models often rely on navigation graphs, 360-degree views, and perfect localization which pose significant challenges when adapting these models to real-world settings. This work addresses these limitations by performing a simulation-to-real domain shift of a VLN approach that operates in continuous environments without requiring navigation graphs or panoramic views. The proposed system integrates vision-language models that align visual inputs and linguistic instructions within a shared embedding space, facilitating natural language-driven navigation. We employ a Cross-Modal Attention (CMA) based architecture trained on an existing dataset in a simulated environment and fine-tune it using real-world data collected from a …

A sense of independence is critical for individuals with mobility impairments, with stair negotiation remaining one of the most demanding daily tasks. This article presents a novel track-based stair-climbing mechanism inspired by leech locomotion, designed to enhance adaptability to varying staircase geometries. The mechanism emulates leech-like motion during both ascent and descent, decomposing each maneuver into multiple segments to ensure stable posture and safety. Throughout motion, the mechanism maintains a large contact surface with the stairs, improving stability. Simulation and experimental studies validate the effectiveness of the proposed motion, particularly during transitions at the top and bottom of the staircase. Results demonstrate that the motion profile reduces abrupt movements and improves user comfort during stair-climbing operation.