How can the President of the United States promise more jobs in terms of percentages or numbers without qualifying those numbers by stipulating that robots, and not humans, may fill many of those positions?
As we continue to advance technologically, an increasing number of robots are performing tasks that people once performed. With every new robotic innovation increasing the abilities of — and the consequent applications for — machines in general, there are fewer jobs remaining for biologically challenged human beings.
Considering efficiency factors alone, robots are approaching 100% utilization for on-the-job availability and performance. Aside from its near 0% absenteeism, a mechanical device does not require benefits, counseling for conflict resolution, more than the minimal amount of square footage to operate, or facilities like restrooms, cafeterias, and vending machines; indeed, every square foot of space occupied by these human-serving amenities could be converted to additional space for factory and production. Thus, in addition to the significant work efficiency gain, robotic technology yields tremendous space efficiency gains that can reduce overhead costs associated with building maintenance, leasing fees, and insurance premiums — all based on square footage.
Cost concerns exist throughout the supply chain, so wherever productivity can be increased for the same cost and time, the business conclusion is straightforward: If robots can do the job, then robots should do the job. Now we can modernize the much-overused quip, “The best man for the job is a woman” to “The best man for the job is a robot.” And with that, we have just identified two additional areas of improvement for the Human Resources department: no gender discrimination claims, and no ageism lawsuits. The robotic factory has pretty strong appeal from many management perspectives.
If I were a robot manufacturer and I wanted to sell to a company that required highly repetitive processes for production operations, I would study the movements performed by the humans doing the job and then determine how many of those movements and decisions I would be able to incorporate into a machine. I would not want my basic robot to be customized to just one company's product line, however; instead, I would expand my research to include fundamental practices across a large industry sector. The price and feature set of my introductory-level robot would cover 80% to 90% of all the tasks required.
In other words, I would want to boast that my robot was the most “universal” in all its capabilities and could replace the greatest number of human workers with a single purchase. That is what causes investors' ears to open and their interest to peak.
At present, the possibilities for 1:1 robot-for-man replacements are not a reality. For some time to come, there will be human jobs, such as design engineering, that will focus creative effort on designing faster, more dexterous, cheaper, more reliable robots — the end goal being to replace even more jobs now held by humans. Ultimately, the engineer's objective is to design the ideal factory, which would produce 100% reliable and identical products at the least possible total cost and in the least amount of time. That means removing all the variables that contribute to inconsistency, workmanship errors, and factory line-downs. To that last point, robot maintenance will become the last bastion of human involvement on the factory floor.
Let me present the factory of the future. By using both emerging and converging technologies, our ideal factory is soon to be located just around the corner of time and innovation. As stated earlier, maintenance becomes the major issue with factory productivity once all other efficiencies have been achieved. Downtime for robots will delay production unless the delay can be compensated for either by maintaining redundant operations or by retaining 100% backup robot reserves that can be brought online without significant interruption to the factory build workstation's operation. Having two identical robots side by side as a fail-over solution may minimize downtime, but it would be more costly from a capital equipment investment and space utilization perspective.
If a robot's failure mechanism and timing can be adequately anticipated, however, then an emerging technology known as “additive manufacturing” can come into play. Additive manufacturing technology is an innovation whereby an object can be produced with equipment that uses the object's original computer-aided design files to build up or print the object in 3D, layer by layer. Just as an inkjet printer reads the computer language that interprets a document for 2D printing, the 3D printer reads the CAD data and prints or manufacturers a solid part.
A reliability engineer (RE) can calculate and anticipate a single component or system's usable lifetime via testing and analysis. For instance, when an RE can gather actual performance and failure-mode data for a specific gear, running 24/7, that is made of a certain material experiencing stresses from temperature and friction at a given speed, then he or she can begin to predict failure rates in and over time. To be more specific, gear G527-000-01 in robot X51 at station W27 performing work instruction WI-19-2 will show sufficient wear after 150,000 hours such that replacement will be mandated within the next 2,000 hours of operation.
In other words, if failures can be anticipated, then robot availability on the production line can be maximized (assuming replacement items are available). If every component has a calculated failure rate, then a spare-parts inventory can be made readily available through materials planning disciplines. Thus, with additive manufacturing, you can go one better.
Taking it one step further, if there are 3D printers in the factory that are connected to a centralized computer system containing all of the CAD, materials and reliability data for every part of every robot, then the mainframe computer could monitor each robot's reliability “health” in real-time and initiate 3D printing activities for the replacement parts just before a failure occurs. The part's plastic materials or metal-alloy powders could be robotically prepared and loaded into bins with RFID location tags that would allow a robot picker to retrieve and set up the printing machines with the correct materials for producing gear number G527-000-01 just hours before it is needed.
With a bank of these printers, repairs based on replacement parts couldn't be more efficient — unless the factory also had a mobile repair robot team or stationary seven-axis, telescoping robot service arms centrally positioned in clusters of production line robots. The service robot would perform no function other than replacing parts and maintaining robots using vision information technology for guiding the arms in the discrete movements required for actions such as fastener and part removal and replacement. These repair robots would have auto-switchable tools at the end of each arm that would make them multifunctional. With perfect, computer-controlled movements in 3D space, the almost-out-of-commission robots would only have to stop long enough to have their local robotic repair arms perform the programmed and precise surgical movements needed to extend their operating life.
With cameras for eyes, along with motion-sensing capabilities, sensors for such variables as temperature and actuator status, accelerometers for 3D space self-awareness, and very fine resolution stepper motors for precise movements, a robot equipped to do everything it needs to do at its assigned workstation need not look like a human. A typical factory robot is bolted to the floor or mounted on rails with shock absorbers to minimize extraneous, vibration-induced movements that promote friction wear, fretting corrosion, and loss of hardware-fastener holding integrity.
Some robotic arms can operate in a 360-degree fashion, swiveling around their base, thereby creating a reach or operation envelope that maximizes their utility by minimizing operational blind spots. If stationary humans could move their arms as effectively as robotic arms sporting capabilities such as telescoping extensions, multi-axis gimballing joints, the ability to lift 10 or 100 times their weight, and the wherewithal to operate without water, sleep, or food, then we would forever remain the superior race. We have mobility that robots do not. But the Defense Advanced Research Projects Agency has engaged project consultants who are busy submitting robot designs and prototypes that can run faster than a cheetah (one is actually called the Cheetah), are more powerful than a steaming locomotive, or, in the case of one fully operational robot called the Sand Flea, are able to leap buildings in a single bound. Sound familiar?
I know the President is a powerful man and can help set policy that will determine the nation's future. But unless he or others in his cabinet have run a comparison analysis of human jobs vs. automation-enabled job replacements over the next five, 10, and 20 years, I don't think he should be raising the public's expectations for long-term job security or new employment opportunities based upon the revival and growth of the manufacturing sector.
I am not discouraging any actions that would bring factory jobs back home, but if we want to play on a level field, we will need to “robot up” quickly to compete with the rest of world. It is a fact that worldwide productivity will continue to rise with the aid of robots. Currently, international productivity is a figure based upon work output per man-hour. Ultimately, with concerns about the environment and the finite supply of fossil fuels, the productivity calculation will be based on output per joule as a measure of energy use and efficiency for work accomplished.
On behalf of myself and other robots here today, I would like to ask Model X51Mobile Rev 15.0 to please come to the podium and accept this year's productivity award.
This article was originally published in EBN's Velocity e-magazine.