Smart as they are, I doubt robots worry about things that keep CEOs up at night. But then, robots don’t “stay up at night,” unless they’re running continuously. Insomnia is just another cultural difference between us humans and our new co-workers, the sort of thing we’d want to talk about over a pizza get-together. Except that robots don’t eat, either.
Since I’ve been reading about automation and robotics, I’ve begun to distinguish the straightforward reporting from the desperate guessing. The first is like steel wire, useful and self-evident. The second is more like silk floss, requiring special handling and prone to unraveling. I’ve seen many examples of each, always attempting to lasso this subject, and they tend to tangle together. Where robotics is concerned, we really do live in interesting times.
On the factual front, robots make great workers, especially for 3D tasks—the dull, dirty, and dangerous. What they lack in lunchroom humor they make up for in tireless and uncomplaining productivity. With the latest improvements in machine vision, force sensing, and speech recognition, they’re ready to tackle jobs that require finicky precision and collaboration. Robots are, in short, poised to join the team.
Even entrenched alarmists like Marshall Brain, teacher, mental tinkerer, and creator of the website How Stuff Works, say that robots should be viewed as a force for good. Nobody really wants to clean toilets or toil in a cornfield under the pitiless sun. (Brain’s worries have more to do with a nation of unemployed who can’t pay their mortgages, never mind buy the high-quality goods robots will be churning out.) Optimistic guesswork maintains that humans will find other, better, more creative things to do if (or when) robots start outperforming us in cleaning, farming, cashiering, building, driving, flying, writing, doctoring, inspecting… the list is rather long.
Whether robots muscle humans totally off the assembly lines and out of the control rooms remains a guess, although that’s drifting toward a more reliable if lukewarm “not anytime soon.” In the meantime—again guesswork, but say for the next three decades or so—we’re stuck with each other.
This shouldn’t be a cause for concern. Last year, virtual magician Marco Tempest gave an amusing presentation at TED addressing humans’ fears about robots and gently introducing us to the inevitable. Another collaborative robot called EDI helped with the sleight of hand:
Given all this, it’s clear that pretty soon we’ll have to stop watching vids about robots and actually start working with the things. So in the satirical spirit of Ambrose Bierce’s The Devil’s Dictionary, here’s a list of words we’ll hear more often going forward, along with their factual—and speculative—definitions.
Accuracy: The measurement of the deviation between the command characteristic and the attained characteristic. Not to be confused with brain fog or muscle fatigue, which are real but separate issues.
Burn-in: A robot-testing procedure where all components are operated continuously for an extended period of time. This is done at early stages to avoid malfunctions after deployment. With human operators, this is often referred to as “burn-out” (see also “overtime”).
Controller: An information processing device whose inputs are the desired and measured position, velocity, or other pertinent variables in a process (see “boss”).
Dead man switch: A quick transfer of workstation personnel, most often following an extended period of overtime, so that production remains constant (see “enabling device”).
Drop delivery: A method of introducing an object to the workplace by gravity. For robots, this is usually by means of a crane; for humans, by means of a manager’s work boot.
Error: The difference between the actual response of a robot and a command issued. Unlike human error, robotic error is generally not construed as insubordination.
Expandability: Being able to add resources to the system, such as memory, larger hard drive, new I/O card, etc. Not generally covered by health insurance plans.
Flexibility: The ability of a robot to perform a variety of different tasks, up to but not including yoga.
Hazardous motion: Unintended or unexpected robot motion that may cause injury (see also “workman’s compensation”).
Industrial robot: A re-programmable multifunctional manipulator designed to move material, parts, tools, or specialized devices through variable programmed motions for the performance of a variety of tasks (see also “employee” and “minimum wage”).
Intelligent robot: A robot that can be programmed to make performance choices contingent on sensory inputs with little or no help from human intervention (see also “downsizing”).
Joint space: A potentially hazardous mindset around work tools, usually detected by testing bodily fluids.
Manipulator: A consultant with stock options in the artificial intelligence industry who can speak knowledgeably about robotics for the purpose of grasping or moving objects, usually contracts or checks (see also “master-slave manipulator”).
Operator: The person designated to start, monitor, and stop the intended productive operation of a robot. An operator may also interface with a robot for productive purposes so long as it’s done off hours.
Presence-sensing safeguarding device: A device designed, constructed, and installed to create a sensing field to detect an intrusion into such field by people, robots, or objects (see “German shepherd”).
Quality assurance (QA): Describes the methods, policies, and procedures necessary to plan, create, and conduct quality assurance testing during the design, manufacturing, and delivery phases of creating, reprogramming, or maintaining robots. Preferably done in triplicate.
Record-playback robot: A manipulator for which the critical points along desired trajectories are stored in sequence by recording the actual values of the joint-position encoders as the robot is moved (see “disk jockey,” “DJ,” and “booking agent”).
Reliability: The probability or percentage of time that a device will function without failure over a specified time period or amount of usage. For a detailed explanation of the phenomenon, see “Murphy’s law.”
Spline: A smooth, continuous function used to approximate a set of functions uniquely defined on a set of subintervals. The approximating function and the set of functions being approximated intersect at a sufficient number of points to ensure a high degree of accuracy in the approximation. So there.
Uptime: A period of time in which a robot is operating or available to operate, as opposed to downtime. Determining uptime is generally expensive and time-consuming, requiring years of transactional analysis (see also “antidepressant”).
Work envelope: A paper pocket or covering to hold items from management and given to employees. Work envelopes originally held wages but were later used to distribute pink slips before being discontinued.
World coordinates: A reference coordinate system in which the manipulator arm moves in linear motions along a set of Cartesian or rectangular axes in X, Y, and Z directions. First proposed in 1978 by the band Village People, who demonstrated it in the hit single, “YMCA”:
Yaw: Rotation of the end-effector in a horizontal plane around the end of the manipulator arm. Also, an informal verbal prompt used to set a robot in motion, sometimes expressed as “ye-haw” or “giddyup.”
Comments
Thanks for the fun break
I enjoyed the light-hearted change of pace. Keep up the good work, Taran.
Add new comment