arXiv subjects

David Douglas

Publications and source records attributed to David Douglas.

10 recordsLinked to original sources

arXiv2024

Resolving Ethics Trade-offs in Implementing Responsible AI

While the operationalisation of high-level AI ethics principles into practical AI/ML systems has made progress, there is still a theory-practice gap in managing tensions between the underlying AI ethics aspects. We cover five approaches for addressing the tensions via trade-offs, ranging from rudimentary to complex. The approaches differ in the types of considered context, scope, methods for measuring contexts, and degree of justification. None of the approaches is likely to be appropriate for all organisations, systems, or applications. To address this, we propose a framework which consists of: (i) proactive identification of tensions, (ii) prioritisation and weighting of ethics aspects, (iii) justification and documentation of trade-off decisions. The proposed framework aims to facilitate the implementation of well-rounded AI/ML systems that are appropriate for potential regulatory requirements.

cs.CY↗

arXiv2023

Implementing Responsible AI: Tensions and Trade-Offs Between Ethics Aspects

Many sets of ethics principles for responsible AI have been proposed to allay concerns about misuse and abuse of AI/ML systems. The underlying aspects of such sets of principles include privacy, accuracy, fairness, robustness, explainability, and transparency. However, there are potential tensions between these aspects that pose difficulties for AI/ML developers seeking to follow these principles. For example, increasing the accuracy of an AI/ML system may reduce its explainability. As part of the ongoing effort to operationalise the principles into practice, in this work we compile and discuss a catalogue of 10 notable tensions, trade-offs and other interactions between the underlying aspects. We primarily focus on two-sided interactions, drawing on support spread across a diverse literature. This catalogue can be helpful in raising awareness of the possible interactions between aspects of ethics principles, as well as facilitating well-supported judgements by the designers and developers of AI/ML systems.

cs.CY↗

arXiv2022

Towards Implementing Responsible AI

As the deployment of artificial intelligence (AI) is changing many fields and industries, there are concerns about AI systems making decisions and recommendations without adequately considering various ethical aspects, such as accountability, reliability, transparency, explainability, contestability, privacy, and fairness. While many sets of AI ethics principles have been recently proposed that acknowledge these concerns, such principles are high-level and do not provide tangible advice on how to develop ethical and responsible AI systems. To gain insight on the possible implementation of the principles, we conducted an empirical investigation involving semi-structured interviews with a cohort of AI practitioners. The salient findings cover four aspects of AI system design and development, adapting processes used in software engineering: (i) high-level view, (ii) requirements engineering, (iii) design and implementation, (iv) deployment and operation.

cs.CY↗

arXiv2021

AI Ethics Principles in Practice: Perspectives of Designers and Developers

As consensus across the various published AI ethics principles is approached, a gap remains between high-level principles and practical techniques that can be readily adopted to design and develop responsible AI systems. We examine the practices and experiences of researchers and engineers from Australia's national scientific research agency (CSIRO), who are involved in designing and developing AI systems for many application areas. Semi-structured interviews were used to examine how the practices of the participants relate to and align with a set of high-level AI ethics principles proposed by the Australian Government. The principles comprise: (1) privacy protection and security, (2) reliability and safety, (3) transparency and explainability, (4) fairness, (5) contestability, (6) accountability, (7) human-centred values, (8) human, social and environmental wellbeing. Discussions on the gained insights from the interviews include various tensions and trade-offs between the principles, and provide suggestions for implementing each high-level principle. We also present suggestions aiming to enhance associated support mechanisms.

cs.CY↗

arXiv2021

Software Engineering for Responsible AI: An Empirical Study and Operationalised Patterns

Although artificial intelligence (AI) is solving real-world challenges and transforming industries, there are serious concerns about its ability to behave and make decisions in a responsible way. Many AI ethics principles and guidelines for responsible AI have been recently issued by governments, organisations, and enterprises. However, these AI ethics principles and guidelines are typically high-level and do not provide concrete guidance on how to design and develop responsible AI systems. To address this shortcoming, we first present an empirical study where we interviewed 21 scientists and engineers to understand the practitioners' perceptions on AI ethics principles and their implementation. We then propose a template that enables AI ethics principles to be operationalised in the form of concrete patterns and suggest a list of patterns using the newly created template. These patterns provide concrete, operationalised guidance that facilitate the development of responsible AI systems.

cs.AI↗

arXiv2015

Linear Vlasov solver for microbunching gain estimation with inclusion of CSR, LSC and linac geometric impedances

As is known, microbunching instability (MBI) has been one of the most challenging issues in designs of magnetic chicanes for short-wavelength free-electron lasers or linear colliders, as well as those of transport lines for recirculating or energy recovery linac machines. To more accurately quantify MBI in a single-pass system and for more complete analyses, we further extend and continue to increase the capabilities of our previously developed linear Vlasov solver [1] to incorporate more relevant impedance models into the code, including transient and steady-state free-space and/or shielding coherent synchrotron radiation (CSR) impedances, the longitudinal space charge (LSC) impedances, and the linac geometric impedances with extension of the existing formulation to include beam acceleration [2]. Then, we directly solve the linearized Vlasov equation numerically for microbunching gain amplification factor. In this study we apply this code to a beamline lattice of transport arc [3] following an upstream linac section. The resultant gain functions and spectra are presented here, and some results are compared with particle tracking simulation by ELEGANT [4]. We also discuss some underlying physics with inclusion of these collective effects and the limitation of the existing formulation. It is anticipated that this more thorough analysis can further improve the understanding of MBI mechanisms and shed light on how to suppress or compensate MBI effects in lattice designs.

physics.acc-ph↗

arXiv2015

Multistage CSR microbunching gain development in transport or recirculation arcs

Coherent synchrotron radiation (CSR) induced microbunching instability has been one of the most challenging issues in the design of modern accelerators. A linear Vlasov solver has been developed [1] and applied to investigate the physical processes of microbunching gain amplification for several example lattices [2]. In this paper, by further extending the concept of stage gain as proposed by Huang and Kim [3], we develop a method to characterize the microbunching development in terms of stage orders that allow the quantitative comparison of optics impacts on microbunching gain for different lattices. We find that the microbunching instability in our demonstrated arcs has a distinguishing feature of multistage amplification (e.g, up to 6th stage amplification for our example transport arcs, in contrast to two-stage amplification for a typical 4-dipole bunch compressor chicane). We also try to connect lattice optics pattern with the obtained stage gain functions by a physical interpretation. This Vlasov analysis is validated by ELEGANT [4] tracking results with excellent agreement.

physics.acc-ph↗

arXiv2015

CSR induced microbunching gain estimation including transient effects in transport and recirculation arcs

The coherent synchrotron radiation (CSR) of a high brightness electron beam traversing a series of dipoles, such as transport or recirculation arcs, may result in the microbunching instability (μBI). To accurately quantify the direct consequence of this effect, we further extend our previously developed semi-analytical Vlasov solver [C. -Y. Tsai et al., FEL Conference 2014 (THP022)] to include more relevant coherent radiation models than the steady-state free-space CSR impedance, such as the entrance and exit transient effects derived from upstream beam entering to and exiting from individual dipoles. The resultant microbunching gain functions and spectra for our example lattices are presented and compared with particle tracking simulation. Some underlying physics with inclusion of these effects are also discussed.

physics.acc-ph↗

arXiv2014

Theoretical investigation of coherent synchrotron radiation induced microbunching instability in transport and recirculation arcs

The coherent synchrotron radiation (CSR) of a high brightness electron beam traversing a series of dipoles, such as recirculation or transport arcs, may lead to the microbunching instability. We extend and develop a semi-analytical approach of the CSR-induced microbunching instability for a general lattice, based on the previous formulation with 1-D CSR model [Phys. Rev. ST Accel. Beams 5, 064401 (2002)] and apply it to investigate the physical processes of microbunching amplification for two example transport arc lattices. We find that the microbunching instability in transport arcs has a distinguishing feature of multistage amplification (e.g, up to 6th stage for our example arcs in contrast to two stage amplification for a 3-dipole chicane). By further extending the concept of stage gain as proposed by Huang and Kim [Phys. Rev. ST Accel. Beams 5, 074401 (2002)], we developed a method to quantitatively characterize the microbunching amplification in terms of iterative or staged orders that allows the comparison of optics impacts on microbunching gain for different lattices. The parametric dependencies and Landau damping for our example lattices are also studied. Excellent agreement of the gain functions and spectra from Vlasov analysis with results from ELEGANT is achieved which helps to validate our analyses.

physics.acc-ph↗

arXiv2000

The Jefferson Lab 1 kW IR FEL

The Jefferson Lab (JLab) IR Demo Free Electron Laser (FEL) has completed commissioning and is initiating user service. The FEL - a high repetition rate, low extraction efficiency wiggler-driven optical cavity resonator - produces over 1 kW of tuneable light on intervals in a 3-6 micron wavelength range. It is driven by a 35-48 MeV, 5 mA superconducting RF (SRF) based energy-recovering continuous wave (CW) electron linac. The driver accelerator meets requirements imposed by low energy, high current, and a demand for stringent beam control at the wiggler and during energy recovery. These constraints are driven by the need for six-dimensional phase space management, the existence of deleterious collective phenomena (space charge, wakefields, beam break-up, and coherent synchrotron radiation), and interactions between the FEL and the accelerator RF. We will detail the system design, relate commissioning highlights, and discuss present performance.

physics.acc-ph↗