- 演講時間:114/5/8(四)15:30-17:00
- 演講地點:國立清華大學工程一館107演講廳
- 演講者: 黃晧倫 總經理 (達詳自動化股份有限公司)
- 講題: 從需求出發,開創藍海新事業 (Expert in Smart Manufacturing for the Metal Industry)
本次演講首先將簡介工具機所面臨的技術挑戰,並以五軸工具機為例,說明如何透過數位雙生技術來預測加工精度、加工時間以及解析加工紋路,以提升機台的性能,最後將介紹如何應用AI人工智慧於加工程式生成、控制器參數優化、加工紋路鑑別、刀具磨耗偵測以及溫昇補償等技術。
This presentation will begin by introducing the technology challenges of the machine tool. Then using five-axis machine tools as an example to illustrate how digital twin technology can be employed to predict machining accuracy, machining time, and analyze cutting marks such that the performance of the machine tool can be enhanced significantly. Finally, it will explore the application of AI in areas such as NC program generation, controller parameter optimization, cutting mark identification, tool wear detection, and thermal compensation.
Embodied intelligence bridges artificial intelligence and the physical world through real-world interaction. Intelligence emerges not only from computation but from the interaction between a system's physical body and its environment. As intelligent agents interact with the world, their physical structures can evolve, shaping and enhancing their capabilities.
This perspective has renewed focus in robotics on the role of a robot’s mechanical form in shaping its intelligence. Current robotic research falls into two domains. Rigid robots are known for their high precision and force capacity, and soft robots are known for their compliance, adaptability, and safety in unstructured environments. However, both types face limitations in real-world tasks that demand a balance of strength and flexibility.
This talk introduces reconfigurable robotics, a class of robotic systems capable of dynamically altering physical configurations. Furthermore, the metamorphic mechanisms can switch topologies and degrees of freedom without disassembly, enabling robots to adapt efficiently to diverse tasks and environments with minimal energy cost. By integrating mechanisms intelligence with control intelligence, we edge closer to a new generation of embodied intelligent robots—machines that are intelligent not just in code, but also in structure.
Face restoration in the wild is a challenging task, especially when the input comes from low-quality surveillance systems. This talk introduces a dynamic blind face restoration framework designed to handle open-set cross-resolution face recognition (CRFR) problems.
The method targets key challenges including:
We propose a unified solution that includes:
Theoretical analysis and extensive experiments on benchmark datasets such as SCface, CRLFW, and QMUL-Tinyface demonstrate the effectiveness of the proposed method.
Join us to explore how this framework bridges fidelity, visual quality, and real-world applicability in surveillance scenarios.