IEEE Robotics & Automation Magazine - June 2018 - 57

operational service, the robot works in collaboration with two
workers, one at a time.
Operator Role
The operator performs tasks that are difficult to automate. In
this case, the person's job consists of picking up randomly
placed metal reinforcement plates from a bin. These are then
disposed on a cart placed in front of the robot. Another situation
where the operator's intervention is needed concerns tasks
that require knowledge of the overall process. Here, the car
type on the assembly line dynamically changes. Therefore,
the worker needs to exchange messages with the robot to
inform it about the plates that need to be glued. Gestures are
used to indicate the current car type on the assembly line.
For instance, a two-finger signal, confirmed by a thumbs-up,
is performed by the operator to instruct the robot to glue
metal parts corresponding to the two-door car type.
Acceptance by Operators
In terms of the social acceptance of the robot, the interviews
performed at the end of the project with the operators who
used Walt demonstrated that the robot had been accepted as
part of the team. The robot would be a point of discussion
and jest by the operators. Furthermore, they mentioned that
working with the latest technology actually gave them a sense
of pride. It was a topic that they would gladly discuss at home
and with friends [8].
Conclusions
The ClaXon project investigated the social and technological
aspects of cobots in a manufacturing industry in terms of
human-robot and robot-product interaction. Throughout
the project, several tools were developed that allow faster and
more flexible introduction of industrial and cobots and the
demonstration of complex robot behavior. To assess the social
acceptance of the technologies developed during the project,
experimental studies were performed with factory workers to
study the human-robot interaction and evaluate the importance of social cues.
The project's research results led to the integration of a
cobot for the gluing process of car reinforcement plates in
the Audi Brussels manufacturing production plant. While a
successful cobot was implemented for a real industrial application, the Audi use case corresponds more to a human-
robot cooperation rather than a real human-robot
collaboration application; the latter would imply, for
instance, direct interactions with the robot. This is due to
regulations regarding safety. Although some new standards
have been developed in the context of human-robot collaboration, they are not yet enforced and therefore required by
the regulators to define the liabilities. In Belgium, there is
still no clear guideline or standard for safety inspection
organs. As a result, the robotic systems are designed such
that direct human-robot interaction is limited to prevent
unforeseen situations for which the manufacturer can be
held liable.

We believe that, in the future, a more dynamic approach to
human-robot interaction should be taken, one that is not driven by legislation but rather by the experience of end users supported by technological
advancements. The use of
cobots is gaining traction
The total quantity of glue
for applications within
small- and medium-sized
used during the process,
enterprises, where expensive safety measures and
i.e., the quantity used for a
cages are too big an investment. However, as soon as
determined number of cars,
the applicability of cobots
in our daily tasks becomes
decreased by 20%.
more apparent, cobot
manufacturers will be
forced to give priority to advances in end-user requirements and technology and to drive legislation toward a
clear and acceptable set of safety regulations.
Acknowledgments
ClaXon is a project realized in collaboration with Imec, with
project support from VLAIO (Flanders Innovation and
Entrepreneurship) and Innoviris.
References
[1] F. Cuenca, K. Coninx, D. Vanacken, and K. Luyten, "Graphical toolkits for rapid prototyping of multimodal systems: A survey," Interact.
Comput., vol. 27, no. 4, pp. 470-488, 2014.
[2] F. Cuenca, J. Van den Bergh, K. Luyten, and K. Coninx, "Hasselt:
Rapid prototyping of multimodal interactions with composite eventdriven programming," Int. J. People-Oriented Programming, vol. 5,
no. 1, pp. 19-38, 2016. doi:10.4018/IJPOP.2016010102.
[3] N.Dalal and B. Triggs, "Histograms of oriented gradients for human
detection," in Proc. 2005 IEEE Computer Society Conf. Computer
Vision and Pattern Recognition, June 2005, vol. 1, pp. 886-893.
[4] J. De, D. Vanacken, and C. Raymaekers, K. Coninx, "High-level modeling of multimodal interaction techniques using NiMMiT," J. Virtual
Reality Broadcast., vol. 4, no. 2, 2007.
[5] I. El Makrini, C. Rodriguez-Guerrero, D. Lefeber, and B. Vanderborght, "The variable boundary layer sliding mode control: A safe and
performant control for compliant joint manipulators," IEEE Robot.
Autom. Lett., vol. 2, no. 1, pp. 187-192, 2017.
[6] S. A. Elprama, C. I. C. Jewell, A. Jacobs, I. E Makrini, and B. Vanderborght,
"Attitudes of factory workers towards industrial and collaborative robots,"
in Proc. ACM/IEEE Int. Conf. Human-Robot Interaction, 2017, pp. 113-114.
[7] S. A. Elprama, I. Makrini, V, El, B., and A. Jacobs, "Acceptance of
collaborative robots by factory workers : A pilot study on the importance of social cues of anthropomorphic robots," in Proc. 25th IEEE Int.
Symp. Robot and Human Interactive Communication, 2016, pp. 919-924.
[8] "Special Eurobarometer 427: Autonomous systems," DirectorateGen. Networks, Content, Technol., European Commission, Brussels,
Belgium, June 2015.
[9] M. Fritzsche, N. Elkmann, and E. Schulenburg, "Tactile sensing: A
key technology for safe physical human robot interaction" in Proc. 6th
ACM Int. Conf. Human-Robot Interaction, Mar. 2011, pp. 139-140.

June 2018

IEEE ROBOTICS & AUTOMATION MAGAZINE

IEEE Robotics & Automation Magazine - June 2018

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - June 2018