Australia and New Zealand Banking Group

BackgroundMalalignment of the humeral and glenoid components in total shoulder arthroplasty is associated with complications such as instability, implant loosening, and restricted motion. While the accuracy of patient-specific instrumentation (PSI) for glenoid component orientation is well documented, the precision of PSI for humeral osteotomy has not yet been published for a noncadaveric environment. This study aims to calculate the deviation between the planned and achieved orientations of humeral and glenoid components using PSI in a clinical, in vivo setting.MethodsForty-six patients were enrolled. One patient was excluded due to an intraoperative change to humeral component selection. The gender, etiology for arthroplasty (osteoarthritis or cuff tear arthropathy), severity of glenoid wear based on the Walch classification, and procedure (anatomic or reverse) were recorded. Computed tomography (CT) scans were obtained preoperatively for 3-dimensional reconstruction, from which the target humeral and glenoid component orientations for each patient were selected by the surgeon in the ASTRA surgical planning platform (Enovis, Austin, TX, USA). Humeral and glenoid PSIs were designed and 3-dimensionally printed to execute the planned alignments. A postoperative CT scan was acquired and used to calculate the deviation between the planned and achieved humeral version, humeral neck-shaft angle, humeral osteotomy height, glenoid version, and glenoid inclination. Humeral version was able to be calculated for 22 patients (48.9%) with a CT scan which extended distally beyond the humeral epicondyles.ResultsA total of 8 outliers (17.8%) were identified across all orientation parameters: 5 outliers (11.1%) for humeral orientation and 3 (6.7%) for glenoid orientation. The average deviation was 2.9°, 2.2°, 1.3 mm, 1.7° and 1.4° for humeral version, humeral neck-shaft angle, humeral osteotomy height, glenoid version, and glenoid inclination, respectively. The mean deviation in glenoid version was greater in males than females (P = .007). However, no other statistically significant difference was found between the mean or variance in deviation when comparing cohorts stratified by etiology of arthroplasty, severity of glenoid wear, or surgical procedure.ConclusionPSI generated from CT-based planning can achieve precise execution of preoperative plans in 89% of humeral and 93% of glenoid components in both anatomic and reverse procedures across a variety of patient morphologies. The humeral guide, when used in a more challenging in vivo environment, demonstrates accuracy comparable to those observed in existing ex vivo studies.

Ransomware has grown to be a dominant cybersecurity threat by exfiltrating, encrypting, or destroying valuable user data and causing numerous disruptions to victims. The severity of the ransomware endemic has generated research interest from both the academia and the industry. However, many studies held stereotypical assumptions about ransomware, used unverified, outdated, and limited self-collected ransomware samples, and did not consider government strategies, industry guidelines, or cyber intelligence. We observed that ransomware no longer exists simply as an executable file or limits to encrypting files (data loss); data exfiltration (data breach) is the new norm, espionage is an emerging theme, and the industry is shifting focus from technical advancements to cyber governance and resilience. We created a ransomware innovation adoption curve, critically evaluated 212 academic studies published during 2020 and 2023, and cross-verified them against various government strategies, industry reports, and cyber intelligence on ransomware. We concluded that many studies were becoming irrelevant to the contemporary ransomware reality and called for the redirection of ransomware research to align with the continuous ransomware evolution in the industry. We proposed to address data exfiltration as priority over data encryption, to consider ransomware in a business-practical manner, and recommended research collaboration with the industry.

The increasing popularity of AI, particularly Large Language Models (LLMs), has significantly impacted various domains, including Software Engineering. This study explores the integration of AI tools in software engineering practices within a large organization. We focus on ANZ Bank, which employs over 5000 engineers covering all aspects of the software development life cycle. This paper details an experiment conducted using GitHub Copilot, a notable AI tool, within a controlled environment to evaluate its effectiveness in real-world engineering tasks. Additionally, this paper shares initial findings on the productivity improvements observed after GitHub Copilot was adopted on a large scale, with about 1000 engineers using it. ANZ Bank's six-week experiment with GitHub Copilot included two weeks of preparation and four weeks ofactive testing. The study evaluated participant sentiment and the tool's impact on productivity, code quality, and security. Initially, participants used GitHub Copilot for proposed use-cases, with their feedback gathered through regular surveys. In the second phase, they were divided into Control andCopilot groups, each tackling the same Python challenges, and their experiences were again surveyed. Results showed a notable boost in productivity and code quality with GitHub Copilot, though its impact on code security remained inconclusive. Participant responses were overall positive, confirming GitHub Copilot's effectiveness in large-scale software engineering environments. Early data from 1000 engineers also indicated a significant increase in productivity and job satisfaction.

The recent release of ChatGPT has gained huge attention and discussion worldwide, with responsible AI being a crucial topic of discussion. One key question is how we can ensure that AI systems, like ChatGPT, are developed and adopted in a responsible way? To tackle the responsible AI challenges, various ethical principles have been released by governments, organisations, and companies. However, those principles are very abstract and not practical enough. Further, significant efforts have been put on algorithm-level solutions that only address a narrow set of principles, such as fairness and privacy. To fill the gap, we adopt a pattern-oriented responsible AI engineering approach and build a Responsible AI Pattern Catalogue to operationalise responsible AI from a system perspective. In this article, we first summarise the major challenges in operationalising responsible AI at scale and introduce how we use the Responsible AI Pattern Catalogue to address those challenges. We then examine the risks at each stage of the chatbot development process and recommend pattern-driven mitigations to evaluate the the usefulness of the Responsible AI Pattern Catalogue in a real-world setting.

<p><b>Local governments are innovatively applying smart city technology to resolve challenges in their jurisdictions. These challenges commonly relate to environmental sustainability, infrastructure, and transportation, and result in novel discourses within local government strategies and operations, such as Smart Environment, Smart Infrastructure, and Smart Mobility. Driven by the success of these discourses, local governments seek further solutions through converging the smart city technology with other disciplines. The next likely convergence is with the heritage discipline, subsequently producing the Smart Heritage discourse. Academic literature records that Smart Heritage is an emergent yet unformed discourse that is on the verge of application within local government. Smart Heritage presents opportunities to converge historical narratives with the automated and autonomous capabilities of smart technology. However, due to its novelty, the local government sector has no guidance on delivering Smart Heritage within strategies and operations. Therefore, this thesis comprehensively explores and defines the Smart Heritage discourse and addresses Smart Heritage's delivery within local government strategies and operations.</b></p> <p>The original contributions to knowledge in this thesis are the first thorough definition of Smart Heritage in academic literature and the production of Smart Heritage Principles, which direct the delivery of Smart Heritage within local government. This thesis firstly defines Smart Heritage through an investigation into the nascent patchwork of academic literature at the intersection of the smart city and heritage disciplines. This definition establishes the discursive framework for the subsequent inquiry into how to deliver Smart Heritage in local government organisations. In this inquiry, the researcher conducts three case studies on local governments in Australia: Broken Hill City Council, the City of Melbourne, and the City of Newcastle. In each case study, the researcher analyses strategic smart city and heritage documents and then interviews their smart city and heritage advisors regarding strategic and operational convergences between the disciplines. The researcher then synthesises the resulting data into cross-case key considerations that contextualise a base understanding of how local governments deliver Smart Heritage. Using this understanding, the researcher conducts a second round of interviews and synthesis that, in turn, produces the refined Smart Heritage Principles. The researcher validates the principles’ relevancy and applicability through an additional case study with Wellington City Council in New Zealand.</p> <p>The research finds that Smart Heritage in the academic literature is nascent yet organically forming around a shared discourse between the smart city and heritage disciplines. As a result, there are numerous understandings of Smart Heritage. Nevertheless, these understandings agree that Smart Heritage convergences historical contextual narratives with automatic and autonomous technologies and advances from the passive Digital Heritage discourse. The case studies find that there is a foundation for Smart Heritage within local government through strategic documents that share similar focuses and advisors who seek multi-disciplinary convergences. However, the disciplines’ overlapping is not explicitly recognised in strategic documents and operational models, leading to inadequate financial and staff resourcing of Smart Heritage and inefficient cross-disciplinary initiatives in local government. The research identifies four thematic key considerations that address delivering Smart Heritage within local government; recognition, delivery, resourcing, and innovation; and proposes four Smart Heritage Principles for local governments to follow in order to deliver the discourse. The researcher presents the principles in an industry-ready document at the end of the thesis.</p> <p>The implications of this research are the increased visibility of Smart Heritage as an academic discourse and support for the delivery of Smart Heritage within local government strategies and operations. Smart Heritage becomes more visible as this research solidifies then illuminates a discursive pathway that researchers can engage with. Importantly, this research presents evidence that Smart Heritage is extant within academic literature and local governments, supporting its position as a constructive academic and practical discourse. The Smart Heritage Principles support the delivery of Smart Heritage within local government strategies and operations through the applied guidance they offer the organisations. As the industry-ready document is the first publication with this focus, the influence on the delivery of Smart Heritage is significant. The researcher aspires to share the Smart Heritage Principles document beyond this research context through its distribution to other councils globally.</p>

Hand pose estimation has recently attracted increasing research interest, especially with the advance of deep learning. Albeit many successes, the current state of research does present some opportunities for improvement of estimation accuracy. This paper presents several image enhancement techniques to embed with existing deep learning architectures to improve the performance of hand pose estimation. In particular, we propose a preprocessing approach for image data using a low light illuminance model or a dehazing algorithm before passing the image data to a hand pose estimation model. Both the preprocessing methods are evaluated on a rendered hand-pose dataset using different evaluation metrics. The experimental results shows success in boosting the performance of hand pose estimation for both 2D and 3D image data.

The benchmarking exercise revealed that the key to successfully implementing an online learning system is a well planned and executed implementation strategy. This chapter explores the emergence of online learning in Australia to determine the successful and unsuccessful strategies and approaches to learning. It describes the context of the Australian vocational education and training environment, discusses effective online instructional design and analyses the strategies applied by organisations to implement online learning and the development of content. A key influencer of online learning in Australia is the vocational training system. Research by TMP into the up take of online learning in Australian corporations revealed that training managers expected, on average, that e-leaming would represent 39% of all training conducted in the workplace. The chapter presents a case study exploring the successful strategy used by the Australia and New Zealand Banking Group to implement online learning into their organisation.