Introduction
Populated settings such as festivals, conventions, or emergency evacuations require efficient crowd management and safe information dissemination. Traditional service robots are often rigid, costly, and limited to structured indoor environments. To address these challenges, this work introduces a mobile robot with an inflatable interface that emphasizes safety, adaptability, agility, and affordability.
Methodology
The platform, nicknamed Martha, combines a rover base with a soft, inflatable interface. The rover base uses skid steering, four motors, and a lightweight aluminum chassis, capable of jogging-speed locomotion over uneven terrain. The inflatable component is made of ripstop nylon and integrates a projector, camera, speakers, and microphone for multimodal interaction. The interface can expand to human-scale for engagement or contract to navigate confined spaces. All control runs on a Raspberry Pi 4 with a m…
Results
The robot weighs 4.5 kg, travels at speeds up to 1.71 m/s, and supports over an hour of continuous operation. Its inflatable interface fully inflates in approximately 40 seconds and deflates in 50 seconds, enabling both compact storage and large-scale interaction. Gesture recognition, text-to-speech, and camera-projector-based input allow diverse human-robot interaction modes. The entire system costs approximately $1,200 USD and is composed primarily of off-the-shelf components and 3D-printed parts.
Discussion
Compared with prior inflatable robots designed for specific groups (e.g., Puffy for children with developmental disorders), Martha broadens applicability to general audiences in public and emergency contexts. The inflatable design enhances safety, the modular software and hardware architecture facilitates adaptation, and the affordability promotes replication and mass deployment. The approach demonstrates how soft robotics can merge with mobile platforms to support scalable, interactive human-robot sys…
Conclusion
This research introduces a mobile, inflatable-interface robot designed for robust, safe, and versatile human-robot interaction. Its lightweight, modular, and cost-effective design makes it suitable for both research and real-world applications in dynamic, crowded environments. Future work aims to integrate ROS-based autonomy, enhance gesture recognition, and explore additional mechanical capabilities for resilience and cooperative operation.