Introduction to the Hardware
15
EN/ISO 13849-1 defines a reliability level of any safety components in a machine by performance level in
terms of average probability of dangerous failure per hour. It then attempts to provide a statistical method
to compute this number for safety systems based on failure rates for various system components to
determine the actual Performance Level (PL), which can be compared to the Required Performance
Level (PLr).
Maximum Allowable Forces to Prevent Operator Injury.
ISO/TS 15066 provides detailed force limitations (Appendix C) based on extensive ergonomic testing of
both pressure (force per unit area) and forces on various parts of the human body. The reader is referred
to this document for detail, but a rule of thumb for maximum pressure to avoid injury is less than
200N/cm
2
, and the maximum crushing force against a rigid surface (quasi-static) to avoid injury ranges
from 65N for the face to 200N for less sensitive parts of the body, with 130-150N being a good rule of
thumb for any part of the body other than the face. Maximum free space collision forces (transient forces)
are typically two times the allowable crushing force and therefore typically range from 260-300N.
Note that there are other well established references for force levels that will not cause injury to humans.
These include:
Automotive Power Windows: 135N
(1)
Power Operated Pedestrian Doors: 180N
(2)
Elevator Door Maximum Closing Force: 135N
(3)
Note also that recent studies have shown that it is impact force, rather than moving mass, that determines
whether an unconstrained collision in free space will injure a person. For impact forces with blunt
surfaces with the human maxilla (upper jaw bone) must reach 600N to break the bone. This can require
a velocity of over 2 meters/sec.
4
Additional research for safety for collaborative robots is ongoing. The Institute for Occupational Safety
and Health (IFA) in Germany has surveyed the literature relating to crushing and impact injuries. Figure
2 below summarizes their findings, which has contributed to the current draft 15066 standard. Note in
particular the column for CC, or compression constant, for various parts of the body. This data is useful in
determining the stiffness or compliance for force sensors when taking collision data. If a rigid robot part is
driven into a rigid sensor the forces will be unrealistically high when compared to bumping into a more
compliant human.
A useful number that may be extracted from this data for testing is a compression constant of 75N/mm,
which is consistent with the hand and the face. For collisions, a higher compression constant will
generate higher collision forces. It is interesting to note that while front of the neck has a fairly low impact
force pain threshold of 35N, the neck must be compressed 3.5mm to reach this force, while in the case of
the hand, which has an impact force pain threshold of 180N, the compression distance is less, at 2.4mm
even though the force is much higher.
In considering the design and testing of a robot that meets these “Collaborative” standards, the likelihood
of an impact to a particular area should be considered. The hand is most likely to be pinched in any pinch
points, whereas the skull is less likely to be pinched as most operators that may be extending their hands
into the workspace will be quite wary of getting their heads between a moving robot and a hard surface.
(1) National Highway Traffic Safety Administration, 49 CFR Part 571, [Docket No. NHTSA-2004-19032] RIN
2127-AG36, Federal Motor Vehicle Safety Standards; Power-Operated Window, Partition, and Roof Panel
Systems
(2) ANSI/BHMA A156.10-1999 American National Standard for Power Operated Pedestrian Doors
(3) Department of Public Safety Division 40 Chapter 5 Elevators
(4) Safe Physical Human-Robot Interaction: Measurements, Analysis & New Insights, 2010, Sami Haddadin,
Alin Albu-Schaffer, Gerd Hirzinger, Institute of Robotics and Mechatronics DLR e.V. - German Aerospace
Center, P.O. Box 1116, D-82230 Wessling, Germany
To Next Page
Loading...
