| Colloquium Mr. Kimizuka |
Date: | Tuesday 23rd of July, 2024 |
Time: | 14.30h (20 min presentation and 10 min discussion) |
Room: | C2506 |
Title: | Haptic in-sensor computing device made of CNT-PDMS nanocomposites |
Abstract
The development of flexible and highly sensitive pressure sensors for artificial skin applications has been getting much attention. The development of wearable devices directly attached to the human body will significantly help in medical care, nursing care, and welfare, such as prosthetics and soft robotics. As one of the candidates for pressure sensors, carbon nanotube-polydimethylsiloxane (CNT-PDMS) nanocomposite is widely investigated. CNTPDMS is expected to detect pressure with the same sensitivity as human skin, which can fabricate flexible and sensitive sensors. When a mechanical stimulus is applied, a conductive path is formed in the CNTs, which function as sensors due to the change in resistance. In recent years, it has been shown that the CNT-PDMS nanocomposite's porous structure directly influences sensing behavior [1]. In this study, we aim to develop "in-sensor computing," in which the CNT-PDMS nanocomposite itself recognizes objects.
| Colloquium Mr. Muzitani
|
Date: | Tuesday 23rd of July, 2024 |
Time: | 15.00h (20 min presentation and 10 min discussion) |
Room: | C2506 |
Title: | Task Planning and Knowledge Acquisition Systems for Home Service Robots |
Abstract
Home service robots are attracting attention to support our daily lives. The development of these robots is being accelerated through competitions such as RoboCup. The robots are required to plan and execute tasks based on spoken commands. The robots are also required to acquire knowledge specific to each home environment, including object locations and user preferences. We discuss a command understanding function powered by a large language model and a brain-inspired model designed to acquire and utilize environment-specific knowledge for home service robots.