Recent advances in sensing tech and image recognition empower industrial robots to diversify their applications. With growing complexity in user requirements, demand surges for adaptable, intelligent robots. While past research mainly centered on rigid object grasping, recent trends focus on soft objects. This study focuses on soft objects that are prevalent in both industrial and daily contexts, ranging from sponges and bottles to confectionery and plush toys. The objective is to predict their deformation using machine learning models and then apply these predictions in conjunction with stability calculations from existing models to develop optimal grasping strategies for three-dimensional soft objects.

Traditional cold sensation presentation technology induces a sense of coldness by continuously lowering the skin temperature. However, this can lead to user adaptation to the stimulus, making it difficult to perceive new sensory experiences. Furthermore, there are safety concerns associated with excessive cooling, limiting prolonged usage. To address these issues, this study introduces the ZeroThermal technology, which can provide cold sensation feedback with minimal alteration of skin temperature. This innovative approach leverages the human body's natural sensitivity to rapid temperature changes. The technology combines cold air flow and light sources to rapidly switch between rapid cooling and gradual warming stimuli, eliciting a sensation of coldness while maintaining skin temperature fluctuations close to zero.

Highly responsive and intensity-adjustable feedback technology can provide real-time stimuli that match visual information, creating a highly realistic experience. This technology is applicable in natural human-computer interaction, including virtual reality and remote sensing. Our research focuses on thermal feedback. We introduced the vortex tube effect into haptic devices for the first time, leading to the development of the world's first non-contact cooling method capable of delivering various intensities of cold sensations. Building on this, we designed the world's first non-contact thermal feedback system capable of rapidly conveying both cold and warm sensations. This system is highly responsive and intensity-adjustable, combining our novel cooling method with a heating method based on thermal radiation for non-contact thermal interaction, enabling users to interact with virtual worlds using their bare skin.