Machine learning, social learning and the governance of self-driving cars

Classification of autonomous vehicle crash severity: Solving the problems of imbalanced datasets and small sample size

Only a few researchers have shown how environmental factors and road features relate to Autonomous Vehicle (AV) crash severity levels, and none have focused on the data limitation problems, such as small sample sizes, imbalanced datasets, and high dimensional features. To address these problems, we analyzed an AV crash dataset (2019 to 2021) from the California Department of Motor Vehicles (CA DMV), which included 266 collision reports (51 of those causing injuries). We included external environmental variables by collecting various points of interest (POIs) and roadway features from Open Street Map (OSM) and Data San Francisco (SF). Random Over-Sampling Examples (ROSE) and the Synthetic Minority Over-Sampling Technique (SMOTE) methods were used to balance the dataset and increase the sample size. These two balancing methods were used to expand the dataset and solve the small sample size problem simultaneously. Mutual information, random forest, and XGboost were utilized to address the high dimensional feature and the selection problem caused by including a variety of types of POIs as predictive variables. Because existing studies do not use consistent procedures, we compared the effectiveness of using the feature-selection preprocessing method as the first process to employing the data-balance technique as the first process. Our results showed that AV crash severity levels are related to vehicle manufacturers, vehicle damage level, collision type, vehicle movement, the parties involved in the crash, speed limit, and some types of POIs (areas near transportation, entertainment venues, public places, schools, and medical facilities). Both resampling methods and three data preprocessing methods improved model performance, and the model that used SMOTE and data-balancing first was the best. The results suggest that over-sampling and the feature selection method can improve model prediction performance and define new factors related to AV crash severity levels.

Applying the "SOTEC" framework of sociotechnical risk analysis to the development of an autonomous robot swarm for a public cloakroom

The past decade has seen efforts to develop new forms of autonomous systems with varying applications in different domains, from underwater search and rescue to clinical diagnosis. All of these applications require risk analyses, but such analyses often focus on technical sources of risk without acknowledging its wider systemic and organizational dimensions. In this article, we illustrate this deficit and a way of redressing it by offering a more systematic analysis of the sociotechnical sources of risk in an autonomous system. To this end, the article explores the development, deployment, and operation of an autonomous robot swarm for use in a public cloakroom in light of Macrae's structural, organizational, technological, epistemic, and cultural framework of sociotechnical risk. We argue that this framework provides a useful tool for capturing the complex "nontechnical" dimensions of risk in this domain that might otherwise be overlooked in the more conventional risk analyses that inform regulation and policymaking.

Predictive Value of the Total Bilirubin and CA50 Screened Based on Machine Learning for Recurrence of Bladder Cancer Patients

Purpose

Recurrence is the main factor for poor prognosis of bladder cancer. Therefore, it is necessary to develop new biomarkers to predict the prognosis of bladder cancer. In this study, we used machine learning (ML) methods based on a variety of clinical variables to screen prognostic biomarkers of bladder cancer.

Patients and Methods

A total of 345 bladder cancer patients were participated in this retrospective study and randomly divided into training and testing group. We used five supervised clustering ML algorithms: decision tree (DT), random forest (RF), adaptive boosting (AdaBoost), gradient boosting machine (GBM), and extreme gradient boosting (XGBoost) to obtained prediction information through 34 clinical parameters.

Results

By comparing five ML algorithms, we found that total bilirubin (TBIL) and CA50 had the best performance in predicting the recurrence of bladder cancer. In addition, the combined predictive performance of the two is superior to the performance of any single indicator prediction.

Conclusion

ML technology can evaluate the recurrence of bladder cancer. This study shows that the combination of TBIL and CA50 can improve the prognosis prediction of bladder cancer recurrence, which can help clinicians make decisions and develop personalized treatment strategies.

Artificial Intelligence in the Advanced Diagnosis of Bladder Cancer-Comprehensive Literature Review and Future Advancement

Artificial intelligence is highly regarded as the most promising future technology that will have a great impact on healthcare across all specialties. Its subsets, machine learning, deep learning, and artificial neural networks, are able to automatically learn from massive amounts of data and can improve the prediction algorithms to enhance their performance. This area is still under development, but the latest evidence shows great potential in the diagnosis, prognosis, and treatment of urological diseases, including bladder cancer, which are currently using old prediction tools and historical nomograms. This review focuses on highly significant and comprehensive literature evidence of artificial intelligence in the management of bladder cancer and investigates the near introduction in clinical practice.

The system of autono‑mobility: computer vision and urban complexity-reflections on artificial intelligence at urban scale

Focused on city-scale automation, and using self-driving cars (SDCs) as a case study, this article reflects on the role of AI-and in particular, computer vision systems used for mapping and navigation-as a catalyst for urban transformation. Urban research commonly presents AI and cities as having a one-way cause-and-effect relationship, giving undue weight to AI's impact on cities and overlooking the role of cities in shaping AI. Working at the intersection of data science and social research, this paper aims to counter this trend by exploring the reverse perspective: how do cities affect the development, and expose the present limits, of SDCs? The contribution of this paper is threefold. First, by comparing urban and nonurban environments and thoroughly examining the relationship between computer vision and city-specific sociality and form, it defines machine autonomy/automation as a function of the sociotechnical milieu in which an AI system operates. Second, and related, the paper problematizes the notion of SDCs as autonomous technologies and the role it plays in envisioning contending policy arrangements and technical solutions for achieving full driving automation. Finally, the article offers insight into a materialist and spatialized understanding of AI-namely, not as an abstract quality susceptible to replication within discrete machines, but rather as a distributed property emerging through embodied interactions among a multiplicity of agents (human, non-human, and technological) within/with their environments.

DOMINO: Domain-aware loss for deep learning calibration

Deep learning has achieved the state-of-the-art performance across medical imaging tasks; however, model calibration is often not considered. Uncalibrated models are potentially dangerous in high-risk applications since the user does not know when they will fail. Therefore, this paper proposes a novel domain-aware loss function to calibrate deep learning models. The proposed loss function applies a class-wise penalty based on the similarity between classes within a given target domain. Thus, the approach improves the calibration while also ensuring that the model makes less risky errors even when incorrect. The code for this software is available at https://github.com/lab-smile/DOMINO.

Human-machine coordination in mixed traffic as a problem of Meaningful Human Control

The urban traffic environment is characterized by the presence of a highly differentiated pool of users, including vulnerable ones. This makes vehicle automation particularly difficult to implement, as a safe coordination among those users is hard to achieve in such an open scenario. Different strategies have been proposed to address these coordination issues, but all of them have been found to be costly for they negatively affect a range of human values (e.g. safety, democracy, accountability…). In this paper, we claim that the negative value impacts entailed by each of these strategies can be interpreted as lack of what we call Meaningful Human Control over different parts of a sociotechnical system. We argue that Meaningful Human Control theory provides the conceptual tools to reduce those unwanted consequences, and show how "designing for meaningful human control" constitutes a valid strategy to address coordination issues. Furthermore, we showcase a possible application of this framework in a highly dynamic urban scenario, aiming to safeguard important values such as safety, democracy, individual autonomy, and accountability. Our meaningful human control framework offers a perspective on coordination issues that allows to keep human actors in control while minimizing the active, operational role of the drivers. This approach makes ultimately possible to promote a safe and responsible transition to full automation.

Fear of AI: an inquiry into the adoption of autonomous cars in spite of fear, and a theoretical framework for the study of artificial intelligence technology acceptance

Artificial intelligence (AI) is becoming part of the everyday. During this transition, people’s intention to use AI technologies is still unclear and emotions such as fear are influencing it. In this paper, we focus on autonomous cars to first verify empirically the extent to which people fear AI and then examine the impact that fear has on their intention to use AI-driven vehicles. Our research is based on a systematic survey and it reveals that while individuals are largely afraid of cars that are driven by AI, they are nonetheless willing to adopt this technology as soon as possible. To explain this tension, we extend our analysis beyond just fear and show that people also believe that AI-driven cars will generate many individual, urban and global benefits. Subsequently, we employ our empirical findings as the foundations of a theoretical framework meant to illustrate the main factors that people ponder when they consider the use of AI tech. In addition to offering a comprehensive theoretical framework for the study of AI technology acceptance, this paper provides a nuanced understanding of the tension that exists between the fear and adoption of AI, capturing what exactly people fear and intend to do.

Learning from the Failure of Autonomous and Intelligent Systems: Accidents, Safety, and Sociotechnical Sources of Risk

Efforts to develop autonomous and intelligent systems (AIS) have exploded across a range of settings in recent years, from self-driving cars to medical diagnostic chatbots. These have the potential to bring enormous benefits to society but also have the potential to introduce new-or amplify existing-risks. As these emerging technologies become more widespread, one of the most critical risk management challenges is to ensure that failures of AIS can be rigorously analyzed and understood so that the safety of these systems can be effectively governed and improved. AIS are necessarily developed and deployed within complex human, social, and organizational systems, but to date there has been little systematic examination of the sociotechnical sources of risk and failure in AIS. Accordingly, this article develops a conceptual framework that characterizes key sociotechnical sources of risk in AIS by reanalyzing one of the most publicly reported failures to date: the 2018 fatal crash of Uber's self-driving car. Publicly available investigative reports were systematically analyzed using constant comparative analysis to identify key sources and patterns of sociotechnical risk. Five fundamental domains of sociotechnical risk were conceptualized-structural, organizational, technological, epistemic, and cultural-each indicated by particular patterns of sociotechnical failure. The resulting SOTEC framework of sociotechnical risk in AIS extends existing theories of risk in complex systems and highlights important practical and theoretical implications for managing risk and developing infrastructures of learning in AIS.

Realising Meaningful Human Control Over Automated Driving Systems: A Multidisciplinary Approach

The paper presents a framework to realise "meaningful human control" over Automated Driving Systems. The framework is based on an original synthesis of the results of the multidisciplinary research project "Meaningful Human Control over Automated Driving Systems" lead by a team of engineers, philosophers, and psychologists at Delft University of the Technology from 2017 to 2021. Meaningful human control aims at protecting safety and reducing responsibility gaps. The framework is based on the core assumption that human persons and institutions, not hardware and software and their algorithms, should remain ultimately-though not necessarily directly-in control of, and thus morally responsible for, the potentially dangerous operation of driving in mixed traffic. We propose an Automated Driving System to be under meaningful human control if it behaves according to the relevant reasons of the relevant human actors (tracking), and that any potentially dangerous event can be related to a human actor (tracing). We operationalise the requirements for meaningful human control through multidisciplinary work in philosophy, behavioural psychology and traffic engineering. The tracking condition is operationalised via a proximal scale of reasons and the tracing condition via an evaluation cascade table. We review the implications and requirements for the behaviour and skills of human actors, in particular related to supervisory control and driver education. We show how the evaluation cascade table can be applied in concrete engineering use cases in combination with the definition of core components to expose deficiencies in traceability, thereby avoiding so-called responsibility gaps. Future research directions are proposed to expand the philosophical framework and use cases, supervisory control and driver education, real-world pilots and institutional embedding.

Discovering the Landscape and Evolution of Responsible Research and Innovation (RRI): Science Mapping Based on Bibliometric Analysis

The growing number of papers on Responsible Innovation (RI) and Responsible Research and Innovation (RRI) have shaped the popularity and usefulness of RI and RRI as a technology governance concept. This study reviews and assesses the development of RRI research through a bibliometric analysis of 702 RRI-focused papers and 26,471 secondary references published in the Web of Science Core Collection database between 2006 and 2020. Firstly, the paper provides a broad outline of the field based on annual growth trends, journal distribution, and disciplinary distribution for RRI publications. Secondly, this study reveals the current state of RRI research by identifying influential literature, journals, authors, countries, and institutions. Thirdly, a phased keyword analysis is conducted to determine the stage characteristics of the RRI field. Finally, based on the bibliometric analyses, this study summarises the evolutionary trajectory of RRI and makes recommendations for future research directions. As a complement to the previous qualitative literature review, the paper provides a systematic and dynamic understanding of RRI research.

Responsibility of AI Systems

To support the trustworthiness of AI systems, it is essential to have precise methods to determine what or who is to account for the behaviour, or the outcome, of AI systems. The assignment of responsibility to an AI system is closely related to the identification of individuals or elements that have caused the outcome of the AI system. In this work, we present an overview of approaches that aim at modelling responsibility of AI systems, discuss their advantages and shortcomings to deal with various aspects of the notion of responsibility, and present research gaps and ways forward.

Social implications of autonomous vehicles: a focus on time

The urban environment is increasingly engaging with artificial intelligence, a focus on the automation of urban processes, whether it be singular artefacts or city-wide systems. The impact of such technological innovation on the social dynamics of the urban environment is an ever changing and multi-faceted field of research. In this paper, the space and time defined by the autonomous vehicle is used as a window to view the way in which a shift in urban transport dynamics can impact the temporal experience of an individual. Using the finite window of time created by an autonomous vehicle, a theoretical framework is put forward that seeks to show how contrasting narratives exist regarding the experience of the window of time within the autonomous vehicle. By taking a theoretical approach informed by social theory, the dissolution of barriers between separate spheres of life is explored to highlight the increased commodification of time. In focusing on both the space and time created by the autonomous vehicle this approach seeks to highlight how artificial intelligence can provide a contemporary space in the urban environment while also opening a new window of time. The cognitive dissonance observed when comparing the narratives of autonomous vehicle stakeholders and the historical shift in time use leads to a belief that technology makes the user more free in terms of time. With this paper the autonomous vehicle is shown to be an ideal space and time to view the way in which the use of such technology does not increase free-time, but further dissolves the boundaries between what is and what is not work-time.

Minding the gap(s): public perceptions of AI and socio-technical imaginaries

Deepening and digging into the social side of AI is a novel but emerging requirement within the AI community. Future research should invest in an “AI for people”, going beyond the undoubtedly much-needed efforts into ethics, explainability and responsible AI. The article addresses this challenge by problematizing the discussion around AI shifting the attention to individuals and their awareness, knowledge and emotional response to AI. First, we outline our main argument relative to the need for a socio-technical perspective in the study of AI social implications. Then, we illustrate the main existing narratives of hopes and fears associated with AI and robots. As building blocks of broader “sociotechnical imaginaries”, narratives are powerful tools that shape how society sees, interprets and organizes technology. An original empirical study within the University of Bologna collects the data to examine the levels of awareness, knowledge and emotional response towards AI, revealing interesting insights to be carried on in future research. Replete with exaggerations, both utopian and dystopian narratives are analysed with respect to some relevant socio-demographic variables (gender, generation and competence). Finally, focusing on two issues—the state of AI anxiety and the point of view of non-experts—opens the floor to problematizing the discourse around AI, sustaining the need for a sociological perspective in the field of AI and discussing future comparative research.

Atmospheric conditioning: Airport automation, labour and the COVID-19 pandemic

This paper contributes to debates on human-technologies relations and labour geographies. It thinks through how the adoption of automation is mediated by the conditioning effects of atmospheres in space. Taking the occassion of the COVID-19 pandemic, the paper presents a case of how atmospheres are capable of determining the trajectories of automation, and providing the guiding backdrop for technological (un)development. Drawing on 40 semi-structured interviews conducted with airport labour in Singapore in 2020 at the height of the COVID-19 pandemic, the paper offers an analysis of how airport workers variously and viscerally capitulate to, abandon, and/or desire to collaborate with automation, in ways that are both unstable and atmospherically implicated. To the extent that these affective responses have the potential to change the course of technological and labour futures in airport infrastructures, atmospheres - especially those deliberately advocated by the state and airport management - are also a political force to be reckoned with. The paper concludes with a discussion on how a focus on atmospheres can push geographic research on automation in productive and interesting directions. It views automation not just as a collection of abstract artefacts, but projects constantly subject to the conditioning effects of invisible atmospheres.

AT-BOD: An Adversarial Attack on Fool DNN-Based Blackbox Object Detection Models

Object recognition is a fundamental concept in computer vision. Object detection models have recently played a vital role in various applications, including real-time and safety-critical systems such as camera surveillance and self-driving cars. Scientific research has proven that object detection models are prone to adversarial attacks. Although several proposed methods exist throughout the literature, they either target white-box models or specific-task black-box models and do not generalize on other detectors. In this paper, we proposed a new adversarial attack against Blackbox-based object detectors called AT-BOD. The proposed AT-BOD model can fool the single-stage and multi-stage detectors, where we used an optimization algorithm to generate adversarial examples depending only on the detector predictions. AT-BOD model works in two diverse ways, reducing the confidence score and misleading the model to make the wrong decision or hide the object detection models. Our solution achieved a fooling rate of 97% and a false negative increase of 99% on the YOLOv3 detector, and a fooling rate of 61% false-negative increase of 57% on the Faster R-CNN detector. The detection accuracy of YOLOv3 and Faster R-CNN under AT-BOD was dramatically reduced and reached ≤1% and ≤3%, respectively.

Adopting Safe-by-Design in Science and Engineering Academia: The Soil May Need Tilling

In recent years, Safe-by-Design (SbD) has been launched as a concept that supports science and engineering such that a broad conception of safety is embraced and structurally embedded. The present study explores the extent to which academics in a distinctively relevant subset of science and engineering disciplines are receptive towards the work and teaching practices SbD would arguably imply. Through 29 interviews with researchers in nanotechnology, biotechnology and chemical engineering differences in perceptions of safety, life-cycle thinking and responsibility for safety were explored. Results indicate that although safety is perceived as a paramount topic in scientific practice, its meaning is rigorously demarcated, marking out safety within the work environment. In effect, this creates a limited perceived role responsibility vis-à-vis safety in the production of knowledge and in teaching, with negligible critical consideration of research's downstream impacts. This is at odds with the adoption of a broader conception of, and responsibility for, safety. The considerations supporting the perceived boundaries demarcating scientific practice are scrutinized. This study suggests that implementing SbD in academia requires systemic changes, the development of new methods, and attention for researchers' and innovators' elementary views on the meaning of and responsibility for safety throughout the innovation chain.

The politics of scaling

A fixation on 'scaling up' has captured current innovation discourses and, with it, political and economic life at large. Perhaps most visible in the rise of platform technologies, big data and concerns about a new era of monopolies, scalability thinking has also permeated public policy in the search for solutions to 'grand societal challenges', 'mission-oriented innovation' or transformations through experimental 'living labs'. In this paper, we explore this scalability zeitgeist as a key ordering logic of current initiatives in innovation and public policy. We are interested in how the explicit preoccupation with scalability reconfigures political and economic power by invading problem diagnoses and normative understandings of how society and social change function. The paper explores three empirical sites - platform technologies, living labs and experimental development economics - to analyze how scalability thinking is rationalized and operationalized. We suggest that social analysis of science and technology needs to come to terms with the 'politics of scaling' as a powerful corollary of the 'politics of technology', lest we accept the permanent absence from key sites where decisions about the future are made. We focus in on three constitutive elements of the politics of scaling: solutionism, experimentalism and future-oriented valuation. Our analysis seeks to expand our vocabulary for understanding and questioning current modes of innovation that increasingly value scaling as an end in itself, and to open up new spaces for alternative trajectories of social transformation.

Applying attention-based BiLSTM and technical indicators in the design and performance analysis of stock trading strategies

With the development of the Internet, information on the stock market has gradually become transparent, and stock information is easy to obtain. For investors, investment performance depends on the amount of capital and effective trading strategies. The analysis tool commonly used by investors and securities analysts is technical analysis (TA). Technical analysis is the study of past and current financial market information, and a large amount of statistical data is used to predict price trends and determine trading strategies. Technical indicators (TIs) are a type of technical analysis that summarizes possible future trends of stock prices based on historical statistical data to assist investors in making decisions. The stock price trend is a typical time series data with special characteristics such as trend, seasonality, and periodicity. In recent years, time series deep neural networks (DNNs) have demonstrated their powerful performance in machine translation, speech processing, and natural language processing fields. This research proposes the concept of attention-based BiLSTM (AttBiLSTM) applied to trading strategy design and verified the effectiveness of a variety of TIs, including stochastic oscillator, RSI, BIAS, W%R, and MACD. This research also proposes two trading strategies that suitable for DNN, combining with TIs and verifying their effectiveness. The main contributions of this research are as follows: (1) As our best knowledge, this is the first research to propose the concept of applying TIs to the LSTM-attention time series model for stock price prediction. (2) This study introduces five well-known TIs, which reached a maximum of 68.83% in the accuracy of stock trend prediction. (3) This research introduces the concept of exporting the probability of the deep model to the trading strategy. On the backtest of TPE0050, the experimental results reached the highest return on investment of 42.74%. (4) This research concludes from an empirical point of view that technical analysis combined with time series deep neural network has significant effects in stock price prediction and return on investment.

Imaginaries of Road Transport Automation in Finnish Governance Culture—A Critical Discourse Analysis

As transport automation technology continues to emerge, there is a need to engage in the questions of its governing—to find a balance between unreflective enablement and rigid control. An increasing body of literature has begun to address the topic, but only a few studies have examined discourse and culture as central components of the related governance processes. This article aims to analyse the discourse surrounding self-driving vehicles in the Finnish context by drawing from the concept of sociotechnical imaginaries. The critical discourse analysis framework is applied to study a comprehensive set of documents published by Finnish national-level governmental bodies from 2013 to 2020. The analysis identifies four imagined ways of implementing self-driving vehicles into the Finnish transport system and a large set of mostly positive anticipated implications. Moreover, the analysis illustrates the transport automation imaginary’s cultural and spatial detachment, most obvious in the lack of detail and the disconnection between the imagined implementations and the anticipated implications. The findings are convergent with findings from other governance contexts, where discourse has been largely characterised by an unjustified optimism and strong determinism related to the wedlock with the automobility regime. If left unaddressed, such lack of reflectivity will not just lead to a plethora of undesired implications for Finnish society at large but will also signify a failure in developing an adaptive governance culture needed to face challenges of the 21st century.

The attachments of 'autonomous' vehicles

The ideal of the self-driving car replaces an error-prone human with an infallible, artificially intelligent driver. This narrative of autonomy promises liberation from the downsides of automobility, even if that means taking control away from autonomous, free-moving individuals. We look behind this narrative to understand the attachments that so-called 'autonomous' vehicles (AVs) are likely to have to the world. Drawing on 50 interviews with AV developers, researchers and other stakeholders, we explore the social and technological attachments that stakeholders see inside the vehicle, on the road and with the wider world. These range from software and hardware to the behaviours of other road users and the material, social and economic infrastructure that supports driving and self-driving. We describe how innovators understand, engage with or seek to escape from these attachments in three categories: 'brute force', which sees attachments as problems to be solved with more data, 'solve the world one place at a time', which sees attachments as limits on the technology's reach and 'reduce the complexity of the space', which sees attachments as solutions to the problems encountered by technology developers. Understanding attachments provides a powerful way to anticipate various possible constitutions for the technology.

Applying artificial intelligence for cancer immunotherapy

Artificial intelligence (AI) is a general term that refers to the use of a machine to imitate intelligent behavior for performing complex tasks with minimal human intervention, such as machine learning; this technology is revolutionizing and reshaping medicine. AI has considerable potential to perfect health-care systems in areas such as diagnostics, risk analysis, health information administration, lifestyle supervision, and virtual health assistance. In terms of immunotherapy, AI has been applied to the prediction of immunotherapy responses based on immune signatures, medical imaging and histological analysis. These features could also be highly useful in the management of cancer immunotherapy given their ever-increasing performance in improving diagnostic accuracy, optimizing treatment planning, predicting outcomes of care and reducing human resource costs. In this review, we present the details of AI and the current progression and state of the art in employing AI for cancer immunotherapy. Furthermore, we discuss the challenges, opportunities and corresponding strategies in applying the technology for widespread clinical deployment. Finally, we summarize the impact of AI on cancer immunotherapy and provide our perspectives about underlying applications of AI in the future.

Drum breach: Operational temporalities, error politics and WIPP's kitty litter nuclear waste accident

In February 2014 at the WIPP transuranic waste repository in New Mexico, a drum erupted in fire. It exposed 22 people to radiation, shut down the underground facility for 35 months and cost the United States over a billion dollars. Heat and pressure had built up in the drum due to chemical reactions with an organic kitty litter, Swheat Scoop, which had been mistakenly added to it at Los Alamos National Laboratory, the birthplace of the atomic bomb. This article disrupts two prominent narratives: (a) that the accident was induced by a typographical error made after a waste packaging operations supervisor misheard 'inorganic kitty litter' as 'an organic kitty litter' during a meeting, and (b) that it was induced primarily by 'mismanagement' at WIPP, Los Alamos and the DOE's New Mexico field offices. It does so by exploring how a series of overambitious political initiatives, fraught labor relationships, financialized subcontracting arrangements and US Department of Energy (DOE) performance incentives set the stage for Los Alamos's notorious error by accelerating US waste packaging, shipping and repository emplacement rates beyond systemic capacity. Attention to operational temporalities shows how an often-overlooked nexus of schedule pressures, political-economic imperatives and regulatory breakdowns converged to modulate nuclear waste management workflows and, ultimately, trigger a radiological accident.

The effect of confounding data features on a deep learning algorithm to predict complete coronary occlusion in a retrospective observational setting

Aims

Deep learning (DL) has emerged in recent years as an effective technique in automated ECG analysis.

Methods and results

A retrospective, observational study was designed to assess the feasibility of detecting induced coronary artery occlusion in human subjects earlier than experienced cardiologists using a DL algorithm. A deep convolutional neural network was trained using data from the STAFF III database. The task was to classify ECG samples as showing acute coronary artery occlusion, or no occlusion. Occluded samples were recorded after 60 s of balloon occlusion of a single coronary artery. For the first iteration of the experiment, non-occluded samples were taken from ECGs recorded in a restroom prior to entering theatres. For the second iteration of the experiment, non-occluded samples were taken in the theatre prior to balloon inflation. Results were obtained using a cross-validation approach. In the first iteration of the experiment, the DL model achieved an F1 score of 0.814, which was higher than any of three reviewing cardiologists or STEMI criteria. In the second iteration of the experiment, the DL model achieved an F1 score of 0.533, which is akin to the performance of a random chance classifier.

Conclusion

The dataset was too small for the second model to achieve meaningful performance, despite the use of transfer learning. However, 'data leakage' during the first iteration of the experiment led to falsely high results. This study highlights the risk of DL models leveraging data leaks to produce spurious results.

The ethics of algorithms: key problems and solutions

Research on the ethics of algorithms has grown substantially over the past decade. Alongside the exponential development and application of machine learning algorithms, new ethical problems and solutions relating to their ubiquitous use in society have been proposed. This article builds on a review of the ethics of algorithms published in 2016 (Mittelstadt et al. Big Data Soc 3(2), 2016). The goals are to contribute to the debate on the identification and analysis of the ethical implications of algorithms, to provide an updated analysis of epistemic and normative concerns, and to offer actionable guidance for the governance of the design, development and deployment of algorithms.

Shadowcasting tourism knowledge through media: Self-driving sex cars?

Tourism is central to late-modern life, and tourism research that threatens this centrality is prone to media attention. Framed by sociotechnical transitions theory, we introduce the concept of 'shadowcasting' to show how tourism knowledge disseminated through the media, combined with public comments on its reporting, cast shadows that co-constitute imagined futures. We illustrate shadowcasting through a mixed method approach that demonstrates how media reporting and public comments on a recent paper on autonomous vehicles in tourism emerged and diverged from the original paper. Our findings reveal that issues around sex and terrorism were sensationalised, generating diverse public discourses that challenge linear visions of future transport efficiency. Our concluding discussion indicates other tourism research contexts that are most inclined to shadowcasting.

In AI We Trust: Ethics, Artificial Intelligence, and Reliability

One of the main difficulties in assessing artificial intelligence (AI) is the tendency for people to anthropomorphise it. This becomes particularly problematic when we attach human moral activities to AI. For example, the European Commission's High-level Expert Group on AI (HLEG) have adopted the position that we should establish a relationship of trust with AI and should cultivate trustworthy AI (HLEG AI Ethics guidelines for trustworthy AI, 2019, p. 35). Trust is one of the most important and defining activities in human relationships, so proposing that AI should be trusted, is a very serious claim. This paper will show that AI cannot be something that has the capacity to be trusted according to the most prevalent definitions of trust because it does not possess emotive states or can be held responsible for their actions-requirements of the affective and normative accounts of trust. While AI meets all of the requirements of the rational account of trust, it will be shown that this is not actually a type of trust at all, but is instead, a form of reliance. Ultimately, even complex machines such as AI should not be viewed as trustworthy as this undermines the value of interpersonal trust, anthropomorphises AI, and diverts responsibility from those developing and using them.

The Future of Transportation: Ethical, Legal, Social and Economic Impacts of Self-driving Vehicles in the Year 2025

Self-driving vehicles (SDVs) offer great potential to improve efficiency on roads, reduce traffic accidents, increase productivity, and minimise our environmental impact in the process. However, they have also seen resistance from different groups claiming that they are unsafe, pose a risk of being hacked, will threaten jobs, and increase environmental pollution from increased driving as a result of their convenience. In order to reap the benefits of SDVs, while avoiding some of the many pitfalls, it is important to effectively determine what challenges we will face in the future and what steps need to be taken now to avoid them. The approach taken in this paper is the construction of a likely future (the year 2025), through the process of a policy scenario methodology, if we continue certain trajectories over the coming years. The purpose of this is to articulate issues we currently face and the construction of a foresight analysis of how these may develop in the next 6 years. It will highlight many of the key facilitators and inhibitors behind this change and the societal impacts caused as a result. This paper will synthesise the wide range of ethical, legal, social and economic impacts that may result from SDV use and implementation by 2025, such as issues of autonomy, privacy, liability, security, data protection, and safety. It will conclude with providing steps that we need to take to avoid these pitfalls, while ensuring we reap the benefits that SDVs bring.

Co-existence or displacement: Do street trials of intelligent vehicles test society?

This paper examines recent street tests of autonomous vehicles (AVs) in the UK and makes the case for an experimental approach in the sociology of intelligent technology. In recent years intelligent vehicle testing has moved from the laboratory to the street, raising the question of whether technology trials equally constitute tests of society. To adequately address this question, I argue, we need to move beyond analytic frameworks developed in 1990s Science and Technology Studies, which stipulated "a social deficit" of both intelligent technology and technology testing. This diagnosis no longer provides an effective starting point for sociological analysis, as real-world tests of intelligent technology explicitly seek to bring social phenomena within the remit of technology testing. I propose that we examine instead whether and how the introduction of intelligent vehicles into the street involves the qualification and re-qualification of relations and dynamics between social actors. I develop this proposal through a discussion of a field study of AV street trials in three cities in the UK-London, Milton Keynes, and Coventry. These urban trials were accompanied by the claim that automotive testing on the open road will enable cars to operate in tune with the social environment, and I show how iterations of street testing undo this proposition and compel its reformulation. Current test designs are limited by their narrow conception of sociality in terms of interaction between cars and other road users. They exclude from consideration the relational capacities of vehicles and human road users alike-their ability to co-exist on the open road. I conclude by making the case for methodological innovation in social studies of intelligent technology: by combining social research and design methods, we can re-purpose real-world test environments in order to elucidate social issues and dynamics raised by intelligent vehicles in society by experimental means, and, possibly, test society.

Automated vehicles, big data and public health

In this paper we focus on how automated vehicles can reduce the number of deaths and injuries in accident situations in order to protect public health. This is actually a problem not only of public health and ethics, but also of big data-not only in terms of all the different data that could be used to inform such decisions, but also in the sense of deciding how wide the scope of data should be. We identify three key different types of data, including basic data, advanced data and preference data, provide an ethical analysis of the use of these different types of data and of different ways of prioritizing between pedestrians and passengers, and propose four rules that can help set ethical priorities for ethical data use and decision making by automated vehicles.

Responsibility, rationality, and acceptance: How future users of autonomous driving are constructed in stakeholders' sociotechnical imaginaries

Although autonomous driving is expected to provide a solution for various mobility-related issues, ideas on how the technology will actually unfold are vague. Nevertheless, stakeholders in the field hold expectations about the technology and the future users. With very few exceptions, so far research does not focus on these expectations as social constructions of individuals and publics. In addition, these perceptions play only a minor role in the technology-centered debate. Thus, to bring these perceptions to light and to analyze their implications, we draw on the sociotechnical imaginaries approach to reconstruct stakeholders' views of future users and publics. We perform a qualitative content analysis and show that imaginaries unfold along the themes of responsibility for the process of driving, rationality in decision-making, and acceptance for emerging technologies. We discuss how the themes relate to each other, what role science plays, and what implications follow from the respective stakeholders' views.

Algorithmic Decision-Making in AVs: Understanding Ethical and Technical Concerns for Smart Cities

Autonomous Vehicles (AVs) are increasingly embraced around the world to advance smart mobility and more broadly, smart, and sustainable cities. Algorithms form the basis of decision-making in AVs, allowing them to perform driving tasks autonomously, efficiently, and more safely than human drivers and offering various economic, social, and environmental benefits. However, algorithmic decision-making in AVs can also introduce new issues that create new safety risks and perpetuate discrimination. We identify bias, ethics, and perverse incentives as key ethical issues in the AV algorithms’ decision-making that can create new safety risks and discriminatory outcomes. Technical issues in the AVs’ perception, decision-making and control algorithms, limitations of existing AV testing and verification methods, and cybersecurity vulnerabilities can also undermine the performance of the AV system. This article investigates the ethical and technical concerns surrounding algorithmic decision-making in AVs by exploring how driving decisions can perpetuate discrimination and create new safety risks for the public. We discuss steps taken to address these issues, highlight the existing research gaps and the need to mitigate these issues through the design of AV’s algorithms and of policies and regulations to fully realise AVs’ benefits for smart and sustainable cities.

Ethical governance is essential to building trust in robotics and artificial intelligence systems

This paper explores the question of ethical governance for robotics and artificial intelligence (AI) systems. We outline a roadmap-which links a number of elements, including ethics, standards, regulation, responsible research and innovation, and public engagement-as a framework to guide ethical governance in robotics and AI. We argue that ethical governance is essential to building public trust in robotics and AI, and conclude by proposing five pillars of good ethical governance.This article is part of the theme issue 'Governing artificial intelligence: ethical, legal, and technical opportunities and challenges'.

Algorithmic psychometrics and the scalable subject

Recent public controversies, ranging from the 2014 Facebook 'emotional contagion' study to psychographic data profiling by Cambridge Analytica in the 2016 American presidential election, Brexit referendum and elsewhere, signal watershed moments in which the intersecting trajectories of psychology and computer science have become matters of public concern. The entangled history of these two fields grounds the application of applied psychological techniques to digital technologies, and an investment in applying calculability to human subjectivity. Today, a quantifiable psychological subject position has been translated, via 'big data' sets and algorithmic analysis, into a model subject amenable to classification through digital media platforms. I term this position the 'scalable subject', arguing it has been shaped and made legible by algorithmic psychometrics - a broad set of affordances in digital platforms shaped by psychology and the behavioral sciences. In describing the contours of this 'scalable subject', this paper highlights the urgent need for renewed attention from STS scholars on the psy sciences, and on a computational politics attentive to psychology, emotional expression, and sociality via digital media.

Towards social autonomous vehicles: Efficient collision avoidance scheme using Richardson's arms race model

This paper presents the concept of a social autonomous agent to conceptualize such Autonomous Vehicles (AVs), which interacts with other AVs using social manners similar to human behavior. The presented AVs also have the capability of predicting intentions, i.e. mentalizing and copying the actions of each other, i.e. mirroring. Exploratory Agent Based Modeling (EABM) level of the Cognitive Agent Based Computing (CABC) framework has been utilized to design the proposed social agent. Furthermore, to emulate the functionality of mentalizing and mirroring modules of proposed social agent, a tailored mathematical model of the Richardson’s arms race model has also been presented. The performance of the proposed social agent has been validated at two levels–firstly it has been simulated using NetLogo, a standard agent-based modeling tool and also, at a practical level using a prototype AV. The simulation results have confirmed that the proposed social agent-based collision avoidance strategy is 78.52% more efficient than Random walk based collision avoidance strategy in congested flock-like topologies. Whereas practical results have confirmed that the proposed scheme can avoid rear end and lateral collisions with the efficiency of 99.876% as compared with the IEEE 802.11n-based existing state of the art mirroring neuron-based collision avoidance scheme.