Project homeWiki source code of 2. Socio-Cognitive Engineering
Version 9.1 by Clemente van der Aa on 2023/04/08 17:50
Hide last authors
| author | version | line-number | content |
|---|---|---|---|
 |
8.1 | 1 | This course Socio-Cognitive Engineering tries to combine engineering, cognitive psychology, and social sciences to design and develop technological systems that interact with humans in a socially intelligent way. In this project the methodology proposed by the lectures is used, the structure of the preset chapters of this document follow this methodology. The idea that "From the foundation, a design //specification// is derived that defines “what” the system shall do (function) in a set of use cases (“when”) to bring about a desired effect (i.e., the claim, “why”)." |
| 2 | \\\\[[Figure 1>>url:https://www.frontiersin.org/articles/10.3389/frobt.2019.00118/full#F1]] presents an overview of the Socio-Cognitive Engineering (SCE) methodology, distinguishing the foundation, specification, and evaluation. To establish the //foundation//, i.e., the operational demands, technology and human factors, a selection of established human-computer interaction and human factors methods can be applied, e.g., from the People, Activity, Context and Technology (PACT) analyses ([[Benyon, 2019>>url:https://www.frontiersin.org/articles/10.3389/frobt.2019.00118/full#B4]]) or Cognitive Work Analyses ([[Vicente, 1999>>url:https://www.frontiersin.org/articles/10.3389/frobt.2019.00118/full#B79]]; [[Naikar, 2017>>url:https://www.frontiersin.org/articles/10.3389/frobt.2019.00118/full#B52]]). SCE puts specific emphasis on the identification of expert knowledge and cognitive theories that are relevant and can be formalized for implementation in the human-robot knowledge-base. See, for example, the “situated design rationale” method for formalizing and contextualizing behavior change support techniques of [[Looije et al. (2017)>>url:https://www.frontiersin.org/articles/10.3389/frobt.2019.00118/full#B47]]. | ||
| |
2.1 | 3 | |
| 4 | |||
| |
8.1 | 5 | From the foundation, a design //specification// is derived that defines “what” the system shall do (function) in a set of use cases (“when”) to bring about a desired effect (i.e., the claim, “why”). In the //evaluation//, the claims are tested via prototyping or simulations, in order to validate and refine the foundation and design specification. It is an iterative, incremental development process, aiming at a sound, theoretically and empirically grounded, prototype with a coherent description of its design rationale. Each design-test cycle will advance (a) the prototype, (b) its foundation in the human factors, technology and operational demands, and (c) the design specification. For the building, maintaining and re-using of design knowledge, SCE distinguishes the following development principles. First, creating human-centered AI and robots is viewed as an inter-disciplinary collaborative activity with active stakeholder involvement during the complete development process (cf. [[Riek, 2017>>url:https://www.frontiersin.org/articles/10.3389/frobt.2019.00118/full#B67]]). Second, functional modules are defined and tested incrementally in an iterative refinement process. As learning and adaptation are key characteristics of human-AI systems, this process of iterations should continue during the complete life-cycle of these systems. Third, design decisions are explicitly based on claims analyses, explicating the up-downside trade-offs. Fourth, keeping and sharing the design rationale is key for progress and coherence in the development of AI and social robots. Fifth, a common ontology should be developed and implemented, which defines the core concepts, with their relationships, for human-robot collaboration (e.g., tasks) and communication (e.g., style). |