Automation
A Human Factors Approach to Enhanced Machine Learning in Cars
| Document
|
Author Joseph SMYTH, Stewart BIRRELL, Lech BIREK, Kris KOBYLINSKI, Alex MOUZAKITIS and Paul JENNINGS |
| Abstract Using machine learning techniques, it is possible to learn and subsequently automate certain driver-focused features in consumer vehicles. A human factors approach is taken to review current machine learning systems. Subsequently, it is found that current methods used for machine learning involve long learning times and might not be sufficient to grasp a true understanding of interaction, routine and feature use - a new method is proposed. Issues surrounding trust and acceptance in automation are also explored and recommendations made. |
In Two Minds: Distractive and Protective Effects of Passengers in Automated Vehicles
| Document
|
Author David R. Large, Cath Harvey, Emily Shaw, Sparsh Khandeparker, Gary Burnett, Elizabeth Box |
| Abstract We highlight exemplar behaviours elicited through the observation of driver-passenger interactions in a multiple journey driving simulator study, demonstrating both distractive and protective effects when a front-seat passenger is present during SAE level 3 automated driving. |
Developing an abstraction hierarchy for visual displays in semi-automated vehicles
| Document
|
Author Jediah R. Clark, Neville A. Stanton and Kirsten M. A. Revell |
| Abstract Understanding how to begin the process of designing a complex system can be challenging. Fortunately, human factors experts have a range of methods available for each step of the process from modelling domain constraints and affordances to more refined concept generation. For domains of high risk such as shared-control automated driving, understanding what, why and how to construct the environment around the human to optimise communication could be a major factor in preventing fatal collisions. In this presentation, we describe the process of developing an abstraction hierarchy – the process of mapping out physical objects, processes and values within the domain of semi-automated vehicles and present a section of this analysis exploring the implementation of visual displays for effective driver-automation communication. |
Can Mobility as a Service really contribute to transport sustainability and equity?
| Document
|
Author Rich C. McIlroy and John M. Preston |
| Abstract Mobility as a Service (MaaS) aims to facilitate multi-modal, non-car travel by integrating various forms of transportation services into a single, accessible journey-planning and ticketing platform. Literature on MaaS is growing; however, questions remain concerning the features, requirements, and information elements that potential end users themselves consider most important or useful for customer facing MaaS apps, and how those needs depend on factors like gender, age, and where a person lives. This article reports on an on-going project exploring these issues, giving an overview of results from an employer questionnaire and a large focus group study, and introducing a followup questionnaire that is gathering end-user priorities for MaaS systems and interfaces. |
The Chatty Co-Driver: A Linguistics Approach to Human-Automation-Interaction
| Document
|
Author Alexander ERIKSSON and Neville A STANTON |
| Abstract |
Human Machine Teaming and Human Centred Design
| Document
|
Author Sue Ottley, Jaina Mistry, Kerry Tatlock & Chris Vance |
| Abstract Human Machine Teaming (HuMaT) systems are expected to become increasingly important and common in the UK military domain as a result of changes in both technology and the Battlespace; however the role of the human user is expected to be retained, though it may change. HuMaT systems are sociotechnical systems where needs and behaviours may be emergent and hard to predict from previous systems behaviours, therefore framing formal requirements prior to design can be difficult and sometimes counterproductive. The complex and variable nature of interactions within Human Machine Teams (HMT) means that traditional technology centric design processes do not easily support them, if the best use is to be made of both human and technology actors then a process more focused on user goals and constraints is required. This paper outlines the considerations for applying a Human Centred Design (HCD) process to developing HuMaT systems and how HCD addresses the challenges presented by a technology centric process. |
Streamlining experimental processes using bespoke software
| Document
|
Author Daniel Fay, Neville A Stanton & Aaron PJ Roberts |
| Abstract The Command Teamwork Experimental Test-bed project was a programme of work that evaluated future ways of working in control rooms through a series of command team studies. Each study required a large software stack to record the necessary data. For each study, teams of participants operated stations that required configuration before each scenario. This configuration was performed by giving spoken instructions to participants. While these instructions worked, they had issues that negatively affected the experimental procedure. To alleviate these issues, software was developed to configure the software stack for each participant. This software, the ComTET Laboratory Automation and Management Shell, has yielded benefits to both experimenters and participants. It is put forth that similar results could be achieved in other experiments by practitioners using software to bolster their experimental procedure. |
Automated vehicles as a co-pilot: improving communication during the transfer of control
| Document
|
Author Jediah R Clark, Neville A Stanton & Kirsten M A Revell |
| Abstract Increased automation capabilities create new challenges for automobile designers and manufacturers to address, such as out-of-the-loop performance, mode error, poorly calibrated trust, and breakdowns in communication between agents. A proposed approach to dealing with such issues is to view automation as a co-pilot rather through the implementation of cooperative concepts into automation design. The capabilities for agents to collaborate, coordinate and execute tasks in line with mutual goals and expectations is likely to have a great impact on a number of factors including safety, system efficiency, and trust. The Joint Activity framework proposed by Clark (1996), and applied to automation by Klein et al. (2005), provides a firm theoretical basis in which automation design can be built upon. This draws upon four stages of the handover task in automated vehicles and uses the method of ‘action planning’ to aid in the application of the joint activity framework to handover assistant design. Preliminary themes are discussed, alongside possible practical implications for automated vehicle handover design. |
Human factors and the digital railway: Effecting and managing change through innovation and integration
| Document
|
Author Richard Bye |
| Abstract With an infrastructure operating at full capacity, a testing political climate, and a mix of government, regulatory and public pressures, the GB railway is under more strain and scrutiny than ever before. Currently, more than 4.8 million journeys are made across the network every day, a number which is forecasted to increase by around 40% over the next 20 years. As such, the complex sociotechnical system that makes up Britain’s railway is in a constant state of flux, continually evolving to meet the everchanging demands of today, whilst anticipating the myriad needs of tomorrow. Without proactive interventions, the predicted growth in passenger journeys will erode the resilience of the railway, especially on the busiest parts of the network that are already characterised by an extreme density of train services. Passengers expect, and should receive, a right-time rail service all day, every day. However, minor disruptions frequently lead to congestion and delays, and result in losses to public satisfaction and confidence. Technological interventions to optimise system performance, whilst maintaining a continuous focus on passenger, workforce and public safety, require a robust and coordinated approach from the ergonomics and human factors (human factors) community. Human factors practitioners, in collaboration with frontline rail staff, engineers, system architects and policy makers, can innovate through practice and research to integrate digital railway (DR) technologies, and reduce environmental stressors, whilst maximising the return on investments at every level of the sociotechnical system. This work illustrates the application of systems ergonomics to the delivery of DR technologies within safety-critical work environments. The DR project is focused on deploying new technology to maximise train capacity on the existing infrastructure. To do this requires a comprehensive programme of human factors integration to effect the necessary cultural and organisational changes, and in doing so develop appropriate policies, regulations, standards and plans. |
Optronics: Examining future scopes
| Document
|
Author Kiome A. Pope, Aaron P. J. Roberts, Daniel Fay and Neville A. Stanton |
| Abstract A critical task when completing submarine operations in shallow littoral waters is the collection of visual data for safe navigation. Contemporary submarines complete this task using optronics masts. However, a question remains as to whether advancements afforded by such technologies are being fully utilised. The current work assessed optronics mast usage by expert operators to examine new ways of working to maximise the utility of the technology afforded. |
Modelling Resilient Future Transportation Systems: Integrating automated and active travel modes
| Document
|
Author Katie J. Parnell, Neville A. Stanton & Katie L. Plant |
| Abstract Future trends in public mobility highlight the integration of autonomous technologies to our road transportation systems alongside a recent rise in popularity of active travel and micromobility. The interaction between autonomous vehicles and other vulnerable road users is a complex issue. Ongoing work by the authors aims to understand how our road transportation systems can incorporate these future trends, whilst also maintaining safety and resilience by reducing conflict between modes. |
The Impact of Automatic Train Protection on Ways of Working in Australian Rail: Preliminary Findings from the Driver Perspective
| Document
|
Author Ty YOUNG, Anjum NAWEED |
| Abstract Automatic Train Protection (ATP) systems monitor and control train movements independently of the driver to provide increased safety. However, few studies have investigated the impact that ATP has on the driver. Given ATP uptake, there is a need to fill this research gap. Using interviews, this study investigated the impact of ATP on traditional train driving in the Australian context. Preliminary findings suggest that uses of ATP are multifarious from the driver perspective and raise further questions for study. |
Automated Anomalies – Exploring Experiences Commercial Pilots have with Automation
| Document
|
Author Siddharth Shyamsundar & Dr Nigel Hunt |
| Abstract This exploratory study investigates commercial pilots’ lived experiences with their automated cockpits. Using interpretative phenomenological analysis, a unique dual-impact theory of automation on situational awareness is uncovered. Recommendations focusing on integrating adaptive automation, pilot-in-the loop systems, system transparency, effective autonomy and promoting opportunities for applying manual flying skills are provided. |
Rail Control and Automation Technology: Transitions within the European Rail Traffic Management System
| Document
|
Author Arzoo NAGHIYEV, Sarah SHARPLES, Brendan RYAN, Anthony COPLESTONE and Mike CAREY |
| Abstract The European Rail Traffic Management System (ERTMS) is a control and automation system which is being gradually introduced to rail systems across the world. The introduction of ERTMS has meant an increase in the level of automation in train driving. A mixed methods approach was used to investigate transitions on an ERTMS fitted line. Qualitative data was collected as a part of a semi-structured interview study with ERTMS drivers. Quantitative data was collected as part of a larger real world eye-tracking study with ERTMS drivers. The results provide an initial insight into the effect of transitions with ERTMS on train driver behaviour. |
Developing instructional decals for depositing cash at automated teller machines (ATMs)
| Document
|
Author Phil Day, Maggie McKendry, Andrew WD Smith & Marshall Munro |
| Abstract This paper describes the iterative development of instructional decals by means of expert reviews and user testing to solve a field issue. Findings are presented and conclusions drawn on developing language-free, graphical decals to convey note orientation and positioning. |
Automation and Cyber Security Risks on the Railways – the Human Factors implications
| Document
|
Author Eylem Thron & Shamal Faily |
| Abstract Automation improves rail passenger experience but may reduce cyber resilience because it fails to adequately account for human factors. Preliminary results from a study on signallers and automation confirms this, but judicious use of modelling tools may ensure design for automation considers this. |
Human performance and automated operations: A regulatory perspective
| Document
|
Author Linn Iren Vestly Bergh, Kristian Solheim Teigen & Fredrik Dørum |
| Abstract The petroleum industry is becoming increasingly dependent on digital systems, and the companies have ambitious plans for increased use of digital technology – along the entire value chain. Increased levels of digitalisation present major opportunities for efficiency in the oil and gas industry and can also contribute to enhanced levels of resilience to major accident hazards. At the same time, new risks and uncertainties may be introduced. Based on developments in the industry and society in general, the Norwegian Petroleum Safety Authority (PSA) has in recent years pursued targeted knowledge development related to digitalisation and industrial cyber security. The PSA’s follow-up activities related to digitalisation initiatives in the industry have been based on input and experience from several knowledge development projects. In this paper we will give insight into the main regulatory strategies we have used to follow-up initiatives in the industry, present results from audits on automated drilling operations and discuss the results from the follow-up activities in light of current regulatory development. |
Allocation of Function Method to support future nuclear reactor plant design
| Document
|
Author Dr Arzoo Naghiyev, Mr John Mount, Mr Anthony Rice and Mrs Caroline Sayce |
| Abstract Current Allocation of Function methods require significant levels of judgement and interpretation and there is an opportunity to develop an improved capability for Allocation of Function for the context of the Nuclear Reactor Plant. This paper presents the development and application of an Allocation of Function method that provides a flexible and configurable set of tools which can be selected in accordance to the design stage and project requirements. The Allocation of Function method has been designed to be used in an iterative manner throughout the different stages of design development and used to engage with different engineering teams. The method draws upon existing and well-established HF methods to investigate and capture human-system interactions associated with function delivery. It also focuses on cognitive tasks to ensure introduction of automation continues to provide support to the operator. Particular emphasis is placed on mapping and understanding the cognitive processes employed in function delivery to ensure that all functionality and information requirements are captured in future automation design. The method also informs assigning and selecting a Level of Automation to a function. The Allocation of Function method enables integration with Systems Engineering to trade HF requirements against the engineering requirements for provision of automation. |
A Human Factors approach for analysing highly automated systems
| Document
|
Author Ronald W. McLeod & Nora Balfe |
| Abstract Automation rarely if ever removes people from a system: rather, it changes their role in ways that can sometimes be difficult to predict. Learning from innumerable incidents has shown that organisations involved in developing and introducing highly automated systems must give sufficient attention during the design, development and deployment of automation to the role of people in the system. This paper suggests a structured analysis method that could be used early in the development of potentially any automated system to identify where a focused effort on Human Factors issues is likely to be needed. |
Knowing and not knowing as system design imperatives
| Document
|
Author Robert J. Houghton & Karen Lancaster |
| Abstract We discuss the importance of “not knowing” as a design imperative in digital and automated systems with examples drawn from a range of different settings together with discussion of how this might be responsibly addressed based on analysis using E/HF methods. Reflection is also offered on situations where the temptation to ignorance should not be acted on in design terms - or simply ignored - but embraced as a sensitive heuristic tool for detecting wider system design challenges made salient by digitalisation. |
Trust, Comfort, and Communication: Human Machine Interface Testing with Virtual Reality Robot
 |
Author Conne Mara Bazley & Nikko Van Stolk |
| Abstract Would you feel safe and comfortable working side by side on a task with a robot? Researchers conducted Human Machine Interface (HMI) Testing for a proposed new robot as an additional team member for a warehouse facility. The overall purpose of this project was to explore how to build human/robot trust, robot communication, human expectations from robot behaviour, and how to measure the positive or negative effects relating to trust as we test HMI variables. Because of the size and weight of the robot, Virtual Reality (VR) was used to simulate the warehouse environment to test the VR robot. The researchers created four VR sessions to test the new robot and obtained the reactions and responses of 10 participants. Most participants did not have a significant change in their trust in robots' baseline responses. Participants showed overall trust in robots and their comfort and trust in working with the new proposed robot and the new robot's capabilities. Participant comments about suggested further robot improvements were gathered and accompanied the results. The researchers discovered that the HMI testing for the robot was more about defining the borders of comfort rather than trust. Additionally, researchers discovered to first deal with the psychology of trust and comfort, then concentrate on robot indicators. Additional HMI Testing using VR is planned for the proposed changes for the new robot and future new robots and contemporary design and development features. |