AIM 2018 July 9-12, 2018, Auckland, New Zealand
IEEE/ASME International Conference on Advanced Intelligent Mechatronics

Keynote 1

Smart mechatronics in medicine

J. Geoffrey Chase, Distinguished Professor, University of Canterbury, New Zealand

Abstract: Mechatronic devices can often be viewed as just simple devices, where the applications are well-known and understood, and the elements that make them up, such as digital vision, sensors, and automation, are even more well-understood. We thus expect to see these things in clinical, or other, settings, and, these days, in a wide range of consumer uses for smart devices and systems. Medicine is an area that one would still expect to see significant use of these well-known elements in smart systems. However, medicine is seen as requiring special electronics, sensors and systems, but smart systems are actually in very short supply! There is a huge gap between perception and reality in this case. This talk considers how well-known mechatronic medical devices and fundamental computing are revolutionising some areas of medicine with significant clinical impact by combining them to make smart devices and systems/services. The presentation covers three application areas that highlight different aspects. First, the use of tablet computers and computation are used to control blood sugar levels of intensive care patients around the world, including a cloud interface for easy access to data and quality control auditing. Making dumb devices smart. Second, the use of ultrasonic sensors is being developed to create wearable diagnostics that can audit and diagnose impending hip implant failures, far enough before they occur to save cost on revision surgery. Smart sensing and basic mechanics to make an intelligent mechatronic system. Finally, digital cameras, strobe lights and simple actuation are used in a novel breast cancer screening concept in clinical trials to create an all new way of screening patients that is 5x faster and costs 10x less. This is a simple mechatronic system combined with computing to make a very smart mechatronic medical device. These applications range from doing the same care better to all new medical applications and devices. Each uses simple off the shelf components and systems/software to make a novel mechatronic “smart system/device” to enable more efficient healthcare.

Keynote 2

Ionic Polymer-Metal Composites as a Candidate Underwater Active Material

Kwang J. Kim, NV Energy Professor of Energy and Matter,

Director of Active Materials and Smart Living Laboratory, Department of Mechanical Engineering, University of Nevada, USA

Abstract: Researchers have designed robotic mechanisms to mimic aquatic lifeforms in attempt to exploit the available lessons from nature in marine applications. Ionic polymer-metal composite (IPMC) artificial muscles is a low voltage driven actuator exhibiting large “bending” displacement and operates in an aqueous environment. Thus, they are suited for creating artificial fish-like propulsors that can mimic the undulatory, flapping, and complex motions of fish fins. Conversely, they can be envisioned as sensor when they are subject to mechanical deformation. In this talk, I will present some methods of fabrication for biomimetic devices, specifically that of aquatic life, through 3D printing which can potentially incorporate the soft actuation mechanisms during fabrication and printing. It should be noted that the developed methods of 3D printing which utilize artificial muscles will produce devices which are silent in operation, are low in power consumption, and can be designed to operate in different fluid environments.

Keynote 3

Modelling, Verification, Control and Co-Design of a Wave Energy Converter

Ron J Patton,Professor of School of Engineering and Computer Science,

Faculty of Science and Engineering, University of Hull, UK

Abstract:To achieve efficient power conversion, reliability and survivability in a Wave Energy Conversion (WEC) system, it is necessary to carefully understand the device hydrodynamics, combined wave force estimation and control, to tune the WEC performance into resonance. An investigation into a designed 1/50 scale vertical heaving Point Absorber Wave Energy Converter (PAWEC) implemented in a wave tank at the University of Hull has provided a research focus on aspects of: (i) nonlinear PAWEC dynamics; (ii) causalization and estimation of wave excitation force; (iii) scaled wave tank testing; (iv) PAWEC power maximisation control; (v) hydrodynamic modelling studies. The work involves the valuation of various estimation and adaptive control strategies.

The presentation will outline the strategy of numerical modelling of the PAWEC dynamics, wave tank numerical and experimental verification outline along with a comparison of strategies of (i) reactive control; (ii) latching control; (iii) 3-level tracking control along with redundant approaches for estimating the wave excitation force. The concept of redundancy in measurement/estimation will be described leading to suggestions for good fault tolerant control design.

Co-design between PAWEC geometry and external Power Take-Off (PTO) damping has been investigated via a CFD wave tank study showing how a comparison between three buoy geometries leads to a suitably streamlined device with improved hydrodynamic response and power absorption efficiency and low PTO damping. However, for optimum power conversion the estimation/control must be designed

taking the hydrodynamic behaviour into account through joint optimization and co-design.

Keynote 4

Cloud Robotics: The Cloud-Side Story-- Low-Latency and Reliable Cloud Computing for Robotics

Jie Xu, Professor of School of Computing, University of Leeds, UK

Abstract: This talk will discuss the challenges in re-engineering the Cloud for Robotics. The Cloud may suffer from some difficult problems while applied to Robotics applications, including those of high latency and low reliability. Such problems are typical in a computer system with hundreds of thousands of distributed servers, such as Google’s Cloud datacenter and Alibaba’s Fuxi system. We first investigated the root causes of latency and failures in a Cloud datacenter by analysing real-world datacenter tracelogs, including operational data sets from Google, Alibaba and Adapt (UK), and developed a system model that captures a datacenter’s behavioural characteristics, actual resource consumption, and overall system dependability. The model was then trained extensively using the tracelogs, able to predict the system’s run time behavior in an accurate fashion. Tasks can be now scheduled intelligently based on the behavioural prediction, leading to the efficient execution of an application without a long tail (or a long delay) and at the same time the efficient utilisation of server resources. A new method was also developed for performing rapid but low-cost failover so as to achieve a better degree of reliability in a Cloud datacenter.

About AIM 2018