Project Reference No.: UGC/IDS(C)11/E01/24
Project Title: Dynamic stability monitoring and control of construction tower cranes using Digital Triple AI and IoT (SFU / HKMU, THEi)

Abstract

We propose a new digital triple AI methodology for life cycle monitoring complex structures, taking a construction tower crane as an example. The life cycle includes behaviours like ageing, damage, and fatigue, which the usual digital twin ignores. To generate a great amount of damaged and ageing time history data for a real crane to train the AI model is not feasible and expensive. Digital simulation is ideal but hard to verify if the simulated data are sufficiently real. A digital twin for the crane and a digital twin for the prototype are proposed as a digital triple: crane, prototype and digital models. We collect the physical data of a crane such as its geometry by a drone lidar and material to build a digital finite element model (FEM) to get its fundamental frequencies of lateral, nodding, and torsional modes. With these modes, we build some physical prototypes of a 1:50 scale according to the dynamic similarity principle so that the crane and the prototype have similar fundamental frequencies, length scale ratio, time scale ratio, and force scale (mass scale) ratio. The digital crane is scaled down to build the digital prototype. The operation life cycle factors of ageing, damage, and fatigue are introduced to the prototype twins. Once the prototype twin is verified, we simulate damaged data to train the AI according to the features at the foundation (bolts and welding), the frame (flexure and broken members), the payload (overload and pendulation) and the jab (swing) as functions of the accelerometer time history at the driver cab. The time history has more information than the frequency alone. The time history will be compressed by wavelet transform for AI training similar to the human electrocardiogram (ECG) signals AI training. Using ECG alone, one can classify various heart diseases. After the AI is trained and verified, from the accelerometer signal alone, one can determine the ageing and stability concerning the prototype's features. Using the dynamic similarity principle, we scale up the AI for the physical crane. We have two kinds of AI models in mind: a quasi-static AI model and a dynamic AI model. The main difference is in the feature extractions; the former involves calculating statistical measures, transforming variables, or applying domain-specific knowledge to derive relevant features, and the latter relies on mathematic tools, e.g., wavelet scattering on the non-stationary time series, and isolates the features using diffusion table, etc. Since the natural frequencies of the tower crane are identified, we shall design a variable stiffness-tuned mass damper to reduce the vibration using step putters composed of a linear motor and a rod of variable length controlled by a microcontroller. Vibration control is for the prolongation of the structure life. Once established, the methodology can be applied to other engineering structures. Therefore, this proposal is to establish a new methodology that can be used to build ageing AI models for many other kinds of structures to have short and long terms research and impact. Measures for communication and results dissemination, Gantt chart, and intellectual property rights will be considered. The expected scientific impact of the research and education is emphasized. The project has government and industry partners.

Project Reference No.: UGC/IDS(C)16/P01/24
Project Title: Detection and Degradation of Antibiotics in wastewater effluent - ScreenBiotics (HKMU)

Abstract

Due to the indiscriminate use of antibiotics, the incidence of diseases caused by antibiotic-resistant bacteria (“superbug”) is becoming more prevalent. Superbugs are becoming a global threat. Since 2014 the number of drug-resistant bloodstream infections increased by 35% between 2013 and 2017. In 2018, Public Health England reported the case of a UK man infected with a multidrug-resistant form of gonorrhea. The situation is alarming in developing countries especially in Bangladesh, Myanmar, Vietnam, and Thailand. There were an estimated 1.3 million infection-related deaths, accounting for 12.1% of the total deaths in China 2019. Among them, more than 600,000 deaths were associated with antimicrobial resistance (AMR), including 145,000 deaths attributable to AMR. The top 3 AMR associated deaths were carbapenems-resistance A baumannii (18,143), methicillin-resistance S aureus (16,933) and third-generation cephalosporins-resistance E coli (8032). In Hong Kong, local public hospitals reported an increasing trend of methicillin-resistant Staphylococcus aureus (MRSA) in the past few years. Furthermore, the number of cases of Carbapenemase-producing Enterobacteriaceae (CPE) found in Residential Care Homes for the Elderly (RCHEs) sharply increased by two fold from 2020 to 2021. All these incidences necessitate the importance of detecting the presence of antibiotics and degrading them instantly in addressing the above-mentioned issues. Conventional sewage treatment plants have limited capability in handling antibiotics, which also makes them a significant source of antibiotics in surface water. The situation is also true for Hong Kong, with the observation that waters receiving effluents from sewage treatment works are having higher antibiotic concentrations in general.

Presently, the detection of antibiotics requires the use of liquid chromatography–mass spectrometry (LC-MS) systems. However, these systems cannot be implemented for on-site mass-screening in an open environment for real-time monitoring purposes. Therefore, there is a pressing need for alternative screening methods to identify the presence of antibiotics on-site in real-time, which is currently not available. Considering this, the aim of this proposal is to conduct advanced research into developing translatable technologies for cost-effective and portable systems for the detection and degradation of antibiotics. Exploring portable onsite detection platforms can pave the way towards efficient and reliable antibiotic detection in wastewater effluent, safeguarding public health and promoting a safe marine environment. Developing a cost-effective and portable detection system for antibiotics in wastewater effluent requires meeting various technical requirements. These include achieving low detection limits to comply with regulatory standards, ensuring excellent specificity to distinguish antibiotics from complex wastewater effluent, and adhering to international guidelines. In this regard, we would implement Electro-Molecularly Imprinted Polymers (E-MIPs) that are synthetic materials designed to exhibit molecular recognition properties for specific target molecules, providing both sensitivity and specificity in detection.

For degradation, various methods have been applied to remove antibiotics from water, including adsorption, advanced oxidation processes, biological treatment, and photocatalytic degradation. Among these, photocatalytic degradation is considered as an effective method for reducing or removing antibiotics from the environment, with solar energy driving it as a cheap, green, and sustainable solution. Here, we will utilise additive manufacturing techniques to produce photocatalytic membrane for the degradation of antibiotics.

In short, this project aims to develop cost-effective and portable systems to detect and degrade antibiotics in wastewater effluent, addressing the prevalence of antibiotic-resistant bacteria. It explores E-MIPs for sensitive detection and investigates photocatalytic degradation as a sustainable solution for antibiotic removal from water.