How a service knowledge management system fills in the blanks (Part 2)
on 04-07-201611:59 AM - edited
Connected products sound like they’re new, but they’ve actually existed for a while. Vending machines dial out inventory information via the Internet or cellular networks to report what sales and what needs restocking. Other products, such as jet engines or a machine in a factory, do the same thing on a larger scale by collecting and transmitting via IoT gigabytes of data.
These products using cost-effective sensors send status reports or cries for help. But even with all of this available data, key information is still missing. Using information from the sensors to improve a product’s design is great for the future, but customers need something today to address operational problems and have data that’s smart and actionable now. A service knowledge management system helps with this.
Let’s apply this service knowledge management system concept to something a little more complicated: a pump for a deep sea oil rig. Starting in the pump’s design phase, service engineers use the design to model the pump and then put information about its parts into the service knowledge base. They perform failure analyses to predict when and where failures might occur and how bad they could be. They try to prevent those failures by establishing service cycles. They essentially predict pump performance in the field before the pump ever gets to the field. All industries today perform some form of this analysis to establish this service knowledge management system. They define plans that contain service requirements, procedures and resource estimates for expected service activities.
Based on this analysis, product and service engineers may even make design changes to improve product performance, serviceability or possibly to come up with good, better or best product models for the market. Service engineering will use this data to generate the service Bill of Materials and the specific service plans for each product model.
When product and service engineering have approved a product design for the pump, manufacturing plans it out. The manufacturer builds the pump and records its exact configuration, including its mechanical, electrical, electronic and software parts. The build record or as-built configuration record also includes traceability for lot numbers, serial numbers and suppliers used to assemble the pump. Some of these parts will be sensors that connect via the Internet of Things (IoT) to report readings.
The pump is sold with a service contract and installed on a deep sea oil rig. The pump is now operational pumping 500 barrels per day. Imagine this pump’s communication via the IoT. It transmits temperatures, vibration, power levels, pressure and fluid flow readings. It communicates everything we want to know about its health and work performance as a stream of data.
A service knowledge management system can help keep an oil rig up and running.
All of these data points we’ve been receiving from the pump provide us with the ability to develop trending lines. Whether it’s a monitoring program or a person on the rig reading dashboard reports, someone notices that the vibration readings have started treading outside of predicted or desirable ranges. Temperatures are spiking. Fluid flow is dropping.
One possibility could be that the pump has failing bearings, which have started to wear out a seal. Through its sensors, a smart pump could have sent out a distress call and initiated a shutdown, or an operator reading gages could have noticed something wasn’t right, hit the stop button and called it in. Either way, we know we have a problem, and our goal is to get the pump up and running as quickly as possible. Production time for the pump is revenue. No production time, no revenue. We may have seen the trend if we analyze Big Data rapidly enough, but the failure could also have happened too suddenly to predict.
Big data can't always provide the whole picture.
So the service company has to respond quickly to return the pump to operation. What helps them do that? It’s not Big Data. Big Data told them something was wrong, but that’s all it could tell them. What will help them get the pump up and running again is knowing the pump’s configuration, service history and its status. The service team needs to know which parts to bring to the repair. To know that, they need to know which parts the pump was built with and what parts were changed over the pump’s production life as it was serviced. All of this configuration information is in the service knowledge management system.
Because our team can find the necessary information beyond Big Data, it can quickly prepare to repair the pump. Using the service knowledge management system, our team can also look to see if there are other services that could be done during the outage to reduce future disruptive downtimes or failures for the pump.
While big data and analytics may have provided the call to action, it is the deeper knowledge of the asset contained in the service knowledge management system that provided the necessary information for rapid and accurate preparation and execution of the service event to return the pump to operation and potentially reduce future downtime.
This concludes part two of our series on how to use smart data for smart service. In part three, Steve O’Lear examines how a service knowledge management system can observe trends and apply the system to an automotive industry situation. Stay tuned.
About the author Steve O’Lear has been in the information system industry for more than 35 years. He has held positions in consulting, services management, sales and marketing across computer hardware, timesharing services (cloud), supercomputing and custom information management solutions for various industry segments. He has more recently focused on PDM and PLM, and many of his customer engagements have been in the A&D industry and with discrete manufacturers. He is currently focused on product marketing for solutions related to document management and service lifecycle management. Early in his career, Steve recognized the need for manufacturers to manage product development data and processes more holistically and became involved with the development, implementation and marketing of PDM solutions. He has also recognized this need with PLM and is now promoting the importance of the support phase in the product lifecycle as products become platforms for service for manufacturers.