Model-based design of user interfaces to support situation awareness in maintenance

• Ano de defesa: 2016
• Autor(a) principal: Oliveira, Állan César Moreira de
• Orientador(a): Araujo, Regina Borges de
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação - PPGCC
• Palavras-chave em Português: Consciência situacional; Interface do usuário; Design centrado no usuário; Engenharia cognitiva; Design baseado em modelos
• Palavras-chave em Inglês: Situation awareness; User interface; User-centered design; Cognitive engineering; Model-based design
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: https://repositorio.ufscar.br/handle/20.500.14289/8686

Situation Awareness (SAW) is a cognitive process defined as the perception of elements and events within a time frame, the understanding of their situation and the projection of their status. SAW is a prerequisite for decision-making in dynamic and complex systems and errors in SAW are an acknowledged source of human errors and accidents. Its study is pivotal in many industries, such as aviation, military, oil, gas and rails, and it is being increasingly considered for maintenance, since this activity is deemed critical for every industry. New technologies to support maintenance, such as E-Maintenance, will provide easier access to the desired information to technicians, but the way new technologies lead to improved SAW is influenced by how information is presented in the User Interface (UI), and many UIs for maintenance technicians support their decision-making regarding procedural and technical criteria, but not economical, legal, ethical and political. Therefore, they only allow a partial development of the user SAW, but not the complete comprehension and projection of a situation. These UIs ignore information requirements such as: risks and conditions of the environment; automations; actions and decisions of team members; rules, regulations and policies of enterprises. Therefore, the design and development of UIs to improve SAW in maintenance is compromised by the few solutions in the state of the art for SAW supportive UI design, for model-based design process and for frameworks and reference architectures. Cognizant to this gap, this thesis proposes a solution for the design and development of Situation Awareness support User Interfaces (SASUI) for maintenance work. For that three contributions are proposed: a conceptual framework of Situation Awareness Aspects (FSA) that assists developers in structuring heterogeneous sources of data into a knowledge representation model, to obtain a state oriented view of SAW; a multiagent architecture that instantiates and controls UIs to improve their support of SAW, by using a blend of SAW and UI agents to express the situation (and its projection) of real world entities in the UI; a methodology to create Model-based SAW User Interfaces (MBSAW-UI), in which designers model agents that will assist users in acquiring the SAW necessary for their decision-making process. These contributions follow a Cognitive Engineering approach to guide software developers in the UI design process and also empower domain experts to model their UIs, enabling an End-User Development (EUD) paradigm that facilitate future updates to the system. A study case of a maintenance activity was developed to evaluate these solutions, with two interfaces: a UI designed using MBSAW-UI; a UI designed using solely a Hierarchical Task Analysis. An experiment was performed and showed a 78% increase in SAW with the UI designed to support SAW, which lead to enhanced efficacy (3,85x less errors) and safety (3,87x less errors regarding unsafe behavior).