by Henry L. Jones ll, PhD. SmartSynch
Solomon, ancient king of Israel, is known for his wisdom—so much so that credit for possessing “the wisdom of Solomon” is still one of the greatest compliments a leader or judge can be given. Yet do you ever get the feeling that Solomon might not have been so wise if he had been bombarded with e-mails, phone calls, text messages, a thousand cable channels, and an exponentially growing universe of Web pages? These days we have plenty to think about, but time to think is scarce. The result: a mountain of data, and a thimbleful of wisdom.
The need to absorb all available data to take action is not new–it has been faced by engineers, scientists and military and political leaders since the dawn of data collection. The difference today is the scale of data we are asked to manage. Making the most of an advanced metering infrastructure (AMI) system is an example of these new expectations.
Figure 1 uses a series of funnels to represent the process that distills oceans of data into a single wise action. The first funnel collects data: the ones and zeros, values and descriptors, and the relationships between them that are produced by sensors and processes. They usually need some work just to get ready for further use: removing redundancy, checking for errors and verifying sources when appropriate. The data is considered information once it is given some context: Relevant data is identified and then grouped logically, and irrelevant data is put aside. This newly created set of information is transformed into knowledge when calculations–based on experience, expertise or theory–produce reliable, insightful assessments that may be depended upon. With this knowledge in hand, analysis can begin so that wisdom can be derived, for example, “If X and Y are true, then we should do Z.” The final step is to carry out an action. One should never forget that the reason for collecting all that data was to set in motion something real, useful and smart.
This explanation of the data-to-action process is abstract, so the following three concrete examples should clarify the concepts and how to use them. The first example, from outside our industry, shows how broad the problem is. The other two examples focus on needs within AMI implementations.
Our war in Iraq has introduced regular use of unmanned systems for the first time in the history of world conflict. One common use of unmanned air vehicles (UAVs) is to loiter over sensitive areas for long periods to maintain “persistence” in the battlefield. A UAV may collect terabytes of video data while loitering over a particular strategic intersection. The system then requires software to put this data into context, such as a video display application that overlays the new video onto existing maps. The information displayed may be collected over many hours to determine whether anyone or anything has loitered in the intersection–possibly planting a bomb. The knowledge that the intersection has been clear for many hours may be used to develop wisdom about the situation: the intersection is safe. Thus the terabytes of data are distilled into one bit of wisdom, and a commander can take action to route his convoy through that intersection.
AMI projects, which might seem to be a world apart from military operations, will be exposed to the same issues. Multiple channels of data from millions of meters will produce billions of bytes of data, but to what end? The goal must be to improve the performance of the overall system through wise actions. For example, an AMI system should be able to take meter data and place it into the context of neighborhoods and service transformers, and then produce relevant trends and patterns. This information about localized consumption could then be combined with distribution system information to calculate when an area of the service territory might be approaching a critical base load, creating knowledge that the installed transformer is insufficient for further growth. Because few options exist to limit customer growth, the wise conclusion is that the transformer must be upgraded. The action–to send a crew out with a replacement transformer before customers notice an impact on service–reinforces the usefulness of the AMI effort beyond improved customer billing.
Many AMI system deployments also include a significant investment in monitoring and troubleshooting a new network with millions of endpoints, and thus the challenge of distilling data to an action is relevant to the metering data and data about the AMI network itself. For instance, the network must monitor the link status for millions of connections to the meters (and possibly millions more in-home devices). In many systems, this includes the link route through one or more collection points. AMI network management systems use the context of this link status data to create tools for monitoring and troubleshooting the collection points. For example, this information might be examined to calculate which collectors are being used only by a small number of meters. Knowledge about the size of the communities under each collector could then be used to identify load-balancing concerns or, if the size of the community is smaller than the local topology would indicate, a potential problem at the collector. Using expertise and experience, the troubleshooting tools might conclude that the antenna’s local area network antenna is faulty and make the wise action to send a new collector to be installed. A potential problem, if that collector had been asked to suddenly accommodate a much larger community due to system rerouting, would be averted.
What can be done about this challenge with your current or future AMI system? Because expertise and experience are so important in reducing data to actions, the best system choices are those in which you are part of the largest communities available so that you have many other like-minded individuals trying to find common tools, solve similar problems and refine the same processes.
Technologies that utilize broadly accepted standards are naturals for developing such communities. Instead of taking on every challenge alone, you’ll have message boards, user gatherings and Google searches to help leverage many more sharp minds around the globe.
Consider the ultimate goal of the system–to take wise action—from the start. Look for solutions that take this perspective into account, not simply to load you and your utility up with data, but to help you formulate the information, knowledge, and wisdom required to take smart action.
Perhaps Solomon would say the same thing.
Henry L. Jones II is chief technology officer of SmartSynch. An expert in process automation, Jones was director of information services for a robotics firm in Mountain View, Calif. He received a graduate degree in aeronautics and astronautics from Stanford University and a bachelor’s degree in mechanical engineering from the University of Mississippi. Reach him at firstname.lastname@example.org.