|
|
 |
|
|
|
|
|
Innovations |
 |
|
|
|
|
|
|
|
|
|
The Safe Human-aware Space-Sharing Robot
|
|
Currently, robots have to work behind fences to ensure operator safety because of risks due to high-energy motions and control/software faults. For SMEs - which typically are engaged in small-series production - this poses a significant problem in terms of installation cost and time and lack of interaction and tuning during production. Intrinsically safe robots are claimed to be the cornerstone of future manufacturing concepts such as the space-sharing co-worker for human-robot cooperation. Initial research towards safe robot systems has mainly investigated safety issues of specific robot components. It is recognized that an intrinsically safe robot requires research towards safe mechanics, human motion perception, and safety-conformable layouts and controls. In applications where this is not physically possible, a complementary approach based on active dependable control and human-aware motions is needed.
|
|
|
|
|
|
|
Shop-floor-suitable Devices for intuitive Robot Interaction
Today, the combination of the mass, speed and forces needed for typical manufacturing tasks do not allow direct human-robot interaction at acceptable cost and performance. To achieve the acceptance of robots by SME workers, the main goal is to radically decrease the moving mass of today‘s robots while keeping comparable performance and still reducing cost. Significant research activities are aimed at fulfilling both low-inertia and high-performance criteria, but with limited practical success so far. The consortium has the expertise and ambition to propose and analyze new materials, manufacturing processes, structures and kinematical ideas for high-performance, low-inertia, low-cost robots. Robots that are intrinsically safe by the laws of physics are claimed to be possible, and fulfillment of that goal is clearly measurable. An additional benefit of these robots is decreased driving power, which in turn means lower cost and lower energy consumption (direct physical implication).
A parallel kinematic manipulator (PKM) structure represents a key idea for safe but productive and low-cost robots. Still there will be applications that require the properties of CNC machines or special made hard-automation machines. Similarly, there will also be applications where accessibility (e.g., inside work pieces) is more important than very low cost, or when compliant human interaction is most valuable, which motivates the use of serial structures. The SMErobot™ consortium includes world-class competence to make traditional types of robots light-weight and inherently safe. The RTD work will be carried out along these two complementary directions here presented as two approaches that each will enable new SME-suitable solutions.
|
|
|
|
|
|
|
The safe & productive Robot Working without Fences
State-of-the-art sensors (e.g. laser scanners) protect the robot‘s workspace by ‘virtual curtains’. Usually, a binary signal triggers a full motion stop when an object passes the virtual curtain. For future SME applications, when a human enters the robot workspace, motions should be adapted from fast (productive) speeds to safe, slow and compliant motions. This places severe reliability demands on the detection of human presence: advanced sophisticated sensing is needed for tracking the body motion but also ‘off-the-shelf’ and ‘low-tech’ sensors have to be taken into account in order to reach the economic sustainability of the developed applications that will be mainly addressed towards SMEs. A trade-off needs to be clearly identified in order to develop SME-suitable-robotic applications. Furthermore adaptation of manipulator arm motion to human worker‘s actions is crucial not only for safety reasons but also to increase productivity during cooperative tasks (e.g. for PbD) and-or small batches production. Last but not least safety issues have to be clearly addressed analysing and eventually reviewing existing standards.
|
|
Dansk
Deutsch
English
Italiano
Nederlands
Svenska
Top
Print
