1. Predictive maintenance will require fewer personnel as technologies enable remote monitoring and analysis of equipment data. Wireless sensors will automate data collection and make it available online.
2. Specialists may still need to visit sites for complex issues, but routine monitoring and basic analysis can be done remotely. Data will be archived across similar machines to reuse solutions.
3. The combination of wireless sensors, internal networks, and internet access will enable easy collection and sharing of machine data between frontline analysts and specialists anywhere. This will solve the challenge of connecting to individual machines.
1. Predictive maintenance will require fewer personnel as technologies enable remote monitoring and analysis of equipment data. Wireless sensors will automate data collection and make it available online.
2. Specialists may still need to visit sites for complex issues, but routine monitoring and basic analysis can be done remotely. Data will be archived across similar machines to reuse solutions.
3. The combination of wireless sensors, internal networks, and internet access will enable easy collection and sharing of machine data between frontline analysts and specialists anywhere. This will solve the challenge of connecting to individual machines.
1. Predictive maintenance will require fewer personnel as technologies enable remote monitoring and analysis of equipment data. Wireless sensors will automate data collection and make it available online.
2. Specialists may still need to visit sites for complex issues, but routine monitoring and basic analysis can be done remotely. Data will be archived across similar machines to reuse solutions.
3. The combination of wireless sensors, internal networks, and internet access will enable easy collection and sharing of machine data between frontline analysts and specialists anywhere. This will solve the challenge of connecting to individual machines.
Nelson Baxter, Contributing Editor Some things require immediate atten- following thoughts regarding agriculture to get the data. This will almost surely be tion, examples being a flat tire or a heart and manufacturing. “If we look at past true for periodic data collection, first attack. These things, like a broken ma- stages of development we can see that, for stage problem detection and basic analy- chine, require an immediate response instance, in the case of agriculture, the sis. However, when a problem is com- because they bring everything to an system became more and more efficient. plex, it will still require “boots on the abrupt halt. Other things do not require We still grow incredible quantities of ground,” as the military likes to say. An immediate action; however if not ad- corn, wheat, soybeans and cotton; we just editorial from a few years ago titled “The dressed in time, they can result in severe do it with a small fraction of the person- Perils of Troubleshooting in Cyberspace” consequences. Examples are: not chang- nel that it used to take. The same is true goes into the details of why solving ma- ing the oil in your car or not going to the with mining and manufacturing. Tremen- chinery problems from a computer screen dentist. If either of these situations is not dous amounts of coal are mined and vast can be difficult. Detection and basic taken care of today, tomorrow, next week quantities of goods are produced; it just analysis can be performed remotely, but or even next month there is no immedi- takes far less people to do the work.” As specialists will still have to go on-site to ate problem. It is obvious, however, that was pointed out by the futurist, based solve certain classes of problems. in the long term not performing these ac- upon what has happened to other indus- 3. Data from similar machines will be tions will result in major pain and/or ex- tries we can predict with a fair degree of archived so the wheel will not have to be penses. In the predictive monitoring certainty that the information industry reinvented each time a problem develops. world, these expenses are related to the (of which predictive monitoring is a part) The nuclear industry will probably be the consequential damages that result from a will have to, like every other industry, first to use this approach. For example, failure and the even more significant con- accomplish more with fewer people. if there are 37 nuclear units with the same sequential downtime associated with a In the predictive maintenance arena, design of Containment Spray Pumps or breakdown. part of the efficiency increase will be at- Residual Heat Removal Pumps, then It is frequently a struggle to get manage- tained by getting smarter on how we there are obvious advantages of maintain- ment to stay focused on long term issues monitor things. For instance, should all ing a common data base. Following the such as predictive maintenance. All too equipment in a plant be monitored on a nuclear plants will be small process often the problem de jour gets all the at- monthly basis, as is often done? Most plants that have numerous sites with tention. In the past, this lack of long term likely that is not necessary. Should very identical equipment scattered throughout commitment by upper level management critical equipment be monitored more the world. was offset by personnel on the front lines frequently? This is also probably true. We 4. The combination of wireless trans- taking an active interest in predictive can therefore increase both effectiveness ducers, plant networks and the internet maintenance and thereby supporting it and efficiency by more closely studying will become a powerful force enabling from the ground up instead of the top which equipment to monitor and how the collection and transport of data to the down. Cuts in front line personnel are often to do so. The greatest potential for first line analyst and then on to special- now threatening the bottom up approach. efficiency improvements in predictive ists, if necessary. The problem to date has When the overall picture is evaluated, maintenance, however, lies in the syner- been what the phone people call the what we have in effect is a form of eco- gism of combining the technologies that “dreaded last mile.” This refers to the fact nomic Darwinism occurring. Global com- are either presently available or in devel- that you can set up the trunk and branch petition, massive increases in health care, opment. Wireless sensors coupled with lines of a fiber optic network, but the dif- deregulation and demographic shifts are in-plant networks and accessed by the ficulty lies in the need to go the last mile the predators. It is not too difficult to fig- internet promise to automate the collec- into the millions of homes and busi- ure out who the prey are. All of these fac- tion and logging of data. These data will nesses. The last mile issue in regards to tors are pushing towards reducing the then be available for viewing by analysts predictive maintenance (the numerous work force not only in manufacturing, anywhere there is an internet connection. connections to the bearings on each ma- but now also in its support functions, i.e., The vibration analyst could literally be chine) will be solved by a battery- or scav- predictive maintenance. In the past year, sitting in an airport, at home or in a mo- enged-powered wireless connection be- work force reductions have even hit the tel with a wireless internet connection tween the transducers on the machines nuclear industry that heretofore had been looking at the condition of his critical and the plant network or internet connec- immune to job cuts. machines and sending correspondence to tion. See the May 2002 S&V editorial. With these inevitable cuts in person- operations on whether or not the equip- 5. The approach to predictive mainte- nel, where are we headed in the predict- ment will make it to the next outage. nance will vary from plant to plant. In able maintenance business? It is hard to What then will predictive monitoring large plants with hundreds of machines, predict the future precisely, however, as look like in the future? Based upon the there will probably be a hybrid solution well known futurists have indicated, economic forces that are present, the past that consists of the machines that are vi- there are some things we can do to at least history of other industries and the tech- tal to the operation being wirelessly get an idea. For instance, you cannot pre- nology already available or shortly to monitored on perhaps a daily basis and dict what will happen tomorrow, but you come on line, it will probably look some- the non-vital equipment being manually can predict with a fair degree of certainty what like the following: monitored maybe every six months. This that, based upon past experience, the 1. There will be fewer personnel col- would reduce the amount of manual earth will keep turning, the stock market lecting data and performing analysis. The monitoring to a small percentage of what will go up and down, and there will be previously mentioned combined eco- is being done today and at the same time trouble somewhere in the world. We can nomic forces of global competition, sky- provide better coverage for the vital therefore start to look into the future by rocketing health care costs, deregulation equipment. As costs come down, the studying the past, then looking at the and demographic shifts are so compelling number of machines that can be moni- forces of change and our available op- that this is nearly a certainty. tored wirelessly will increase in these tions in regard to responding to these 2. Data will increasingly be brought to large facilities. Intermediate-size plants forces. One well known futurist gave the the analyst rather than the analyst going that cannot justify onsite personnel will
SOUND AND VIBRATION/MAY 2004 5
need to choose between contracting out a manual service and going to a con- tracted wireless service. Small plants, like those that supply liquid oxygen, per- form specialty chemical processes or have pumping operations that are geo- graphically scattered throughout the country would best be served by entirely going to remote monitoring in order to re- duce the travel time required to go from site to site for manual monitoring. Since these types of plants use nearly identical equipment at dozens of facilities, they will be ideal situations for remote moni- toring with all the data going to a com- mon point so a knowledge base can be acquired and the data mined for use in fine tuning the future prediction of prob- lems. If businesses were more humane and considered the immediate welfare of the workers, then things would stay the same and everyone would remain employed doing what they are doing. That has not happened in agriculture or manufactur- ing and it will not happen in our little world of predictive maintenance. And, in the long term it is probably a good thing. What if 50% of our population were still employed in agriculture? The key to the survival of the analysts and to the com- panies they work for is to recognize what is coming and to be on the forefront of making it happen rather than lamenting about the past.