WORKMATE2018: WORKMATE: the WORKplace is better with intelligent, collaborative robot MATEs Brisbane Convention & Exhibition Venue Brisbane, Australia, May 21, 2018 |
Conference website | https://workmate2018.wordpress.com/ |
Submission link | https://easychair.org/conferences/?conf=workmate2018 |
Submission deadline | April 1, 2018 |
The introduction of the Industry 4.0 paradigm is expected to redefine the nature of shop-floor and intra-logistics environments along many directions, including the role played by robots in the manufacturing process. One of the main objectives considered in Industry 4.0 is an increased customer satisfaction via a high degree of product personalisation and just-in-time delivery. On the one hand, a higher level of flexibility in manufacturing processes is needed to cope with such diversified demands, especially in low-automation tasks, and could lead to high stress levels for human operators. On the other hand, skillful robots working alongside humans can be regarded as a valuable aid to shop-floor operators, who can supervise the robots’ work and intervene whenever needed: robots can be tasked with difficult or otherwise stressful operations.
Human-robot cooperation (HRC) processes in shop-floor environments are a specific form of human-robot interaction (HRI) with at least two important peculiarities:
- the cooperation is targeted towards a well-defined objective (e.g., an assemblage, a unit test, a cabling operation), which must be typically achieved in a short amount of time;
- humans need to be (at least partially) in control: although rooted in the cooperation process, their behaviour could be unpredictable in specific cases, with obvious concerns about their safety; they may not fully understand robot goals; robot actions may not be perceived as appropriate for the peculiar cooperation objectives.
As far as the cooperation process is concerned, two high-level directives must be taken into account: on the one hand, cooperation models enforcing the overall task objectives must be adopted; on the other hand, the robot must be flexible enough to assess the human operator’s stress and fatigue, adapt to her/his actions avoiding a purely reactive approach, and make its intentions clear.
These two directives lead to three functional requirements for a HRC architecture. The robot workmate must be able to:
- recognise human operator intentions, actions or their effects;
- adapt its behaviour on the basis of two elements: human operator stress, fatigue and actions themselves, as well as the cooperation objectives;
- employ action or motion planning techniques allowing for human-aware tuning, and fast action re-planning when needed, e.g., when planned actions or motions cannot be executed for sudden changes in the environment or failing modelling assumptions.
The workshop will target four main topics.
The traditional approach according to which HRC systems are developed targets expert operators. This needs to be re-thought, considering instead common people as the target users who will program and operate the robot. This can be achieved by considering natural and intuitive robot programming approaches and interfaces, which do not require expert knowledge [TOPIC 1].
To favour the acceptability and trust of robots as workmate, it is important to enable a smooth interaction between humans and robots. This implies several aspects: human intention and human action recognition processes are of the utmost importance to adapt to the operator’s actions as they unfold [TOPIC 2]; robot motions and gestures should be not only safe, but also aimed at enhancing human trust, following, e.g., bio-inspired patterns [TOPIC 3]; furthermore, robot behaviors and communication modalities should adapt to the operator’s cognitive and physical capabilities, as well as incipient fatigue and stress, sharing autonomy accordingly and supporting the operator when needed [TOPIC 4].
We will try to identify and debate about key aspects related to the use, the acceptability, the trust and the benefits of HRC systems. WORKMATE will foster discussion around such questions as:
- How can a robot understand operator’s actions, intentions and eventually her/his physical and mental status?
- What should a robot do to make its behavior understandable by the operator, therefore enforcing a sense of trust?
- How much information should the robot provide to the operator? What means of communication should the robot use?
- How much information does the robot need from the operator? What are the best ways to represent such knowledge?
- How can a non-expert operator communicate with the robot in a natural manner?
Submission Guidelines
Prospective participants are welcome to submit full papers (4 to 6 pages), to be presented as posters. Submitted papers must follow the template available at https://workmate2018.wordpress.com/call-for-contributions/.
List of Topics
- Natural cooperation between robots and human operators
- Human motion, intention and status understanding
- Human stress and fatigue detection for affective robotics
- Human trust in human-robot cooperation
- Adaptive robot behavior
- Human-robot communication
- Acceptability of human-robot cooperation systems
Organizing committee
- Valeria Villani, University of Modena and Reggio Emilia, IT
- Lorenzo Sabattini, University of Modena and Reggio Emilia, IT
- Barbara Bruno, Univeristy of Genoa, IT
- Fulvio Mastrogiovanni, Univeristy of Genoa, IT
Invited Speakers
- Kai Arras, Albert-Ludwigs-University of Freiburg, DE
- Norbert Krüger, University of Southern Denmark, DK
- Dongheui Lee, Technical University of Munich, DE
- Il Hong Suh, Hanyang University, ROK
- Sami Haddadin, Leibniz University of Hannover, DE
- Erhan Oztop, Ozyegin University, TR
- Francesco Nori, Italian Institute of Technology, IT
- Katia Sycara, Carnegie Mellon University, USA
- Hae Won Park, Massachusetts Institute of Technology, USA
- Rachid Alami, The National Center for Scientific Research, FR
Venue
The conference will be held in the Brisbane Convention & Exhibition Centre in Brisbane, AUS.
Contact
All questions about submissions should be emailed to Valeria Villani (valeria.villani@unimore.it).