T-CDS Special Issue: Introspective Methods for Reliable Autonomy


IEEE Transactions on Cognitive and Developmental Systems
Special issue on: "Introspective Methods for Reliable Autonomy"
Submission deadline: 28 February 2018


   As humans, understanding our own limitations, failures and shortcomings is a key to improvement and development. This knowledge is crucial for altering our behaviours, e.g. to execute tasks in a more cautious way. Correspondingly, equipping robots with a set of skills that allows them to assess the quality of their sensory data, internal models, used methods etc. will greatly improve their overall performance.
   The problem of introspection, directly or indirectly, relates to other research topics: planning, execution monitoring, active perception and mapping. Accordingly, an improved understanding of introspection in robotics has a direct impact on a large variety of application areas (e.g. search and rescue, intralogistics, assistive robotics).
   The introspection impacts the most the following aspects of robotics system: safety, reliability and the maintenance costs.
   Information on the internal state of the robot is crucial to make decisions if it is safe to execute the assigned mission considering not only the current state of the perceived environment, but also the internal state of the robot.
   Continuous monitoring of the internal state of the robot and automatic assessment can also be used to enhance the maintenance process. Information about the internal state of the robot can be used to estimate the likelihood of potential failure and tailor the efforts to prevent it or to speed up the recovery or repair process by providing detailed information to a human operator or even enable self-repair.
   Introspection takes an active role in the process of preventing of malfunctions of the robotic system and help to speed the repair process up. These two features have a direct impact on the running cost of a robotic system. Preventing unplanned interruptions in the robot operation and shortening the time of the planned interruptions has a direct impact on the cost of robot exploitation.
   It is also important to remember that introspective information is a cornerstone of all methods aiming to robotics self-improvement. It provides information crucial to the learning and development process. In this context, it is possible to draw a parallel between human and robotic system. Assessment of the internal state is important input helping to anticipate if the planned action is feasible for the agent (either human or robot). For a complex system, it is difficult to perform such assessment relying only on the predefined set of rules and conditions.
   Therefore, it is necessary to use learning algorithms which will be able to connect the preexisting internal and external conditions with the outcome of a planned action. In such a configuration, a failure became a crucial element of a learning process of an autonomous system.
   Finally, it is important to emphasis that introspection is a topic which spans across multiple fields. The introspection is originally a human ability. It is recent years when the idea of introspection is also becoming present in the field of robotics.
   Therefore, to obtain a complete picture of the problem of introspection in autonomous systems it is important to have a closer look also at psychological aspect of introspection. Moreover, the impact of introspection in the context of the cognitive science cannot be overlooked.

The primary topic of this issue is to present work on:
How to assess the quality of internal models, methods, sensory data and the hardware used by robots and how to alter their behaviour using this information?

The aim of this special issue is fourfold:
* Survey the state of the art in the field.
* Define open research questions in the field.
* Provide a venue to present the recent developments in the field of introspection.
* Present system papers showing how introspection is integrated and affects the performance of a system.


This special issue is addressed to researchers interested in the development of introspective methods for robust autonomy
across different research areas. We expect to receive submissions relevant for following research fields, but to name a few: Long-term autonomy, safe operation of robots under uncertainty, performance awareness, reliable-aware operation,
cooperative robotics, cognitive and learning robots, developmental robotics, Human-Robot Interaction. Introspection
is broad term covering a set of topics. Topics relevant to this special issue includes, but are not limited to:
* Internal assessment (Map quality assessment, Perception quality assessment, Classification quality assessment)
* Analysis (Failure analysis, Execution monitoring, Meta-reasoning)
* Introspection-related actions (Failure recovery, Reconfigurable robots, Planning with uncertainty)


28 February 2018 - Deadline for manuscript submissions.
15 June 2018 - Notification of authors
15 July 2018 - Deadline for submission of revised manuscripts
15 September 2018 - Final decisions


Manuscripts should be prepared according to the “Information for Authors” of the journal found HERE and
submissions should be made through the IEEE TCDS Manuscript centre at https://mc.manuscriptcentral.com/tcds-ieee selecting the category “SI: Introspective Methods for Reliable Autonomy”.


The IEEE Transactions on Cognitive and Developmental Systems (TCDS) focuses on advances in the study of development and cognition in natural (humans, animals) and artificial (robots, agents) systems. It welcomes contributions from multiple related disciplines including cognitive systems, cognitive robotics, developmental and epigenetic robotics, autonomous and evolutionary robotics, social structures, multi-agent and artificial life systems, computational neuroscience, and developmental psychology. Articles on theoretical, computational, application-oriented, and experimental studies
as well as reviews in these areas are considered.


Yaochu Jin
Department of Computer Science, University of Surrey
Surrey, United Kingdom


Tomasz Piotr Kucner
Centre for Applied Autonomous Sensor Systems, Örebro University
Örebro, Sweden

Sören Schwertfeger
School of Information Science and Technology, ShanghaiTech University
Shanghai, China

Martin Magnusson
Centre for Applied Autonomous Sensor Systems, Örebro University
Örebro, Sweden

Achim J. Lilienthal
Centre for Applied Autonomous Sensor Systems, Örebro University
Örebro, Sweden

Rudolph Triebel
Institute of Robotics and Mechatronics, German Aerospace Center
Oberpfaffenhofen-Weßling, Germany

Easy Links