NO.155 Engineering methods and techniques for interactive systems
May 18 - 21, 2020 (Check-in: May 17, 2020 )
- José C. Campos
- University of Minho, Portugal
- Yoshifumi Kitamura
- Tohoku University, Japan
- David Chemouil
- ONERA DTIS & University of Toulouse, France
Description of the Meeting
The main idea behind the organization of this meeting is the joint observation that, on the one hand, the variety of forms of interactive systems actually used is increasing in many businesses and, on the other hand, the requirements, in particular related to safety and security, put on the said systems are getting more stringent. This stems from the realization that latent anomalies in their design or implementation can lead to use error and potential harm.
For instance, present or near-future aircraft cockpits not only feature classic interaction devices but also multi-touch apparatus, mobile systems, voice-controlled systems, etc. New technologies are paradigms are needed to support developing these new interaction devices in a safe and effi cient manner. Standards such as ARINC 661 (from the avionics domain) aim at standardizing the components and technologies available to build these new user interfaces.
At the same time, the prototype-based iterative design, typical of the human-centred process used in the design and development of such interactive systems, which was for a long time left apart from the stringent safety regulations faced by so-called “critical” systems (such as fly-by-wire), is expected to face the same kind of regulation and then will have to adopt more disciplined development processes. Indeed, standards such as ISO 62366-2 are emerging that define usability engineering process targeted at specific domains, in this case medical devices, with the aim of identifying use-related risks arising from poor user interface design.
Use related issues are already a relevant contributing factor to system failure in application domains such as healthcare, avionics and traffi c control, as illustrated by a recent study at the Food and Drug Administration, where about half of all recalls related to software could be traced back to user interface problems in the devices. Additionally, an interactive system will only be secure if users are aware of the security related tasks they need to perform and do not make dangerous use errors when performing them. All in all, as rich interactive systems are spreading, their development and operation will bear new constraints that call for a reinvention of their engineering methods and techniques. By “method”, we not only mean a language or a set of tools but also processes and heuristics that are justified w.r.t. their purpose. Although not exclusive, a particular emphasis on rigorous and/or formal approaches, which have proven their importance in the field of critical embedded systems for instance, will be promoted. The question to address is then which specific rigorous and/or formal techniques are best suited to assess whether an interactive device satisfies requirements that concern its usability and safety.
For these reasons, the aim of this meeting is to gather experience and new ideas from research in various fields of interactive systems and/or engineering methods to delineate new or strengthened directions for future research. The following topics will be addressed:
- Process development: what is the state of the art for the safe development of interactive systems? How to reconcile the classic prototype-based development of such systems with more stringent processes such as those from the aerospace industry?
- Security and dependability: as the complexity of interactive systems increases, what is the impact of dependability and security requirements on such systems? How to express these requirements? How to assess their fulfillment?
- Requirements engineering: model-based requirements engineering (e.g. agent- or goal-oriented methods such as i*/Tropos and KAOS) is now quite established in the requirements engineering community. But is it a good fit for interactive application engineering or is it necessary to extend or amend it? How to express requirements specific to the field of interaction? How to analyze them?
- Languages: as in any engineering discipline, models and therefore languages provide a mean to state a specification or to evaluate a What is the state of the art for modeling interactive applications? Similarly, are there programming techniques or languages to implement them? What about their (formal) semantics and pragmatics?
- Rapid prototyping: rapid and/or iterative prototyping has been an important way of developing interactive systems, often complemented with end-user studies to guide the directions for im- provements. How to keep the benefits of prototyping while providing stronger assurance on the end-result and its conformance to requirements?
- Hazard analysis: what hazard analysis techniques might be applied to identify potential user interaction design problems from early in the process?
- Verification & validation: as interactive applications play a more and more important role in critical systems, how can we test them thoroughly or check that they are free from hazardous errors or that they satisfy their specification?
- Users: A specificity of interactive systems is that for considering the quality (usability, safety, security,…) if the interaction not only the implemented system, but also the user must be con- sidered. It is the combination of user and system that must be Typical software engineering techniques, however, focus mainly on the system side. What are the best ways to fold considerations about the users (their goals, knowledge, capabilities, …) into the analysis?
To enrich the scope of discussions, we will also address the following topics that should provide an interesting perspective of the problems addressed by the meeting.
- Case studies: are there significant case studies that could be shared and reused by the community to serve as test benches for future research?
- Domain-specific interactive systems: we will welcome reports on specific fields for interactive applications, such as aerospace systems, medical systems or video
- User evaluations, experiments: specialists in user evaluations and user studies will be welcome to report on results regarding the way end users interact with
- Ergonomics, social sciences: finally, we will try to get feedback from researchers in ergonomics, psychology and possibly other such scientific Classically, a specificity of interactive appli- cations is how the end-user is really a fundamental part of the system.
- A main goal of the proposed meeting is to promote collaboration between researchers working on the problems above from different but complementary angles, fostering cross-fertilization. Finally, we mention that once this meeting is completed, we plan to publish a book gathering the main contributions and results.