Don’t Deploy IoT Without Knowledge Management

Don’t Deploy IoT Without Knowledge Management

Don’t Deploy IoT Without Knowledge Management

Don’t Deploy IoT Without Knowledge Management

The Internet of Things (ioT) empowers organizations to apply knowledge management (KM) in the physical world. IoT provides a platform for capturing and managing metadata about physical objects. IoT also requires knowledge management for its success. A good IoT implementation requires descriptions and management of architecture, infrastructure, distribution and maintenance. And since no two IoT projects are alike, KM also offers discipline for sharing and adapting lessons learned across projects and industries.

Organizations deploying IoT should think about knowledge management as a core foundation for their IoT projects with the following objectives:

  • Create a live inventory of machinery, equipment or system (capture what we know about the physical structure of the business at a new level of detail)
  • Populate the inventory list with the current state of the equipment. This may come in many forms including on/off, vibration, thermal, sound, visual, etc. (capture the metadata—static metadata, like serial number, location, make, model, configuration—and realtime metadata like temperature, fluid viscosity, vibration, etc.)
  • Tie incoming data to validate and drive business process models. (Examine what decisions can be better made based on IoT data, such as moving from preventative maintenance to condition-based maintenance in a factory).

Metadata about physical objects

IoT offers the capacity for organizations to capture metadata about physical objects. This means that realtime data about equipment, its location, and its state, can be used to understand a manufacturing or other facility configuration rather than relying on static maps and equipment lists. 

Metadata for physical objects may also arrive from secondary sensors, most likely cameras, that increasingly monitor most public areas. Arrays of small cameras could surround a piece of equipment at a relatively low cost, resulting in continuous monitoring of the environment around the equipment, that could help identify leaks on the floor, track human activity like maintenance, even tampering. Security cameras tied to a deeper IoT strategy would result in much greater context and a wider range of risk analysis scenarios such as continuity based on equipment failure, rather than breakdowns caused by malicious acts.

Managing the things

IoT introduces a huge number of new devices that represent a variety of capabilities from passive sensors for identification, to active sensors for thermal, vibration, night-vision and normal light vision capture, along with active sensors that include control logic. This requires that organizations maintain knowledge about their IoT investments, including:

  • Sensor locations and how they are configured (firmware versions, communications stake versions, any fixed items, like fixed IP addresses).
  • A description of what the “thing” is supposed to do.
  • The technical architecture of the sensor network, including communications protocols, encryption, data transport and storage.

Realtime data for a more responsive application of knowledge

In many cases, procedural or process knowledge, when captured cannot be tested regularly against real-world situations. Because knowledge-based IoT explicitly represents data in relationship to sensors, systems and equipment being monitored, the procedural knowledge gets tested regularly. In equipment monitoring, for example, rules or machine learning interpret incoming data. The interpretation creates alerts or guidance based on the incoming data. If a person taking action on an alert finds the alert to be false, even though the data that triggered it is correct, the fault lies not in the IoT sensor, but in the logic used to interpret its data. That is very much a knowledge management issue, requiring a new version of the logic to repair the faulty alert, and a story to be captured to describe why the change was made.

Because knowledge-based IoT explicitly represents data in relationship to sensors, systems and equipment being monitored, the procedural knowledge gets tested regularly.

Improved decision making

Decision making usually improves as that amount of accurate data increases. Knowledge management helps define what data best serves decision making. KM also guides the description of a system of “things” that informs adaptive decision making. The network of sensors identifies objects, the state of parts of objects (such as the motor in a pump) and provides data for deriving overall context. IoT delivers detail and context in realtime. While both may be available from manual collection and correlation, manual work cannot match the swiftness and iterative power of realtime monitoring. IoT accelerates decision making, and ideally with the increase in data, makes it more precise.

Project management and lessons learned

Every IoT project requires a project plan, and every project plan changes the second it hits the real world. Knowledge management should be used to capture the initial plan, as well as any deviations from the plan that required rethinking or problem-solving. Make the repository of lessons available in a shared repository of choice like Microsoft SharePoint, TheBrain or Atlassian Confluence. What matters most is not the repository but the richness of the data and metadata so people can retrieve the lessons learned. Always include the following metadata:

  • Name of the product and version (of the IoT device)
  • What the IoT device monitors
  • What IoT devices combine to provide data about the equipment or machine or environment
  • Description of intent (what were you trying to do?)
  • Description of the issue (what obstacle to your intent were you trying to overcome?)
  • The date (so people are looking at the most recent solutions first, though this is a backup to the version number which is a better match for problem-solving than date)
  • Resolution (how did you overcome the obstacle?)
  • Location of the originating issue (this is important as different facilities have their own configuration, or fall under different regulatory constraints, so a lesson learned in one location may prove unavailable to a team in a different country).

Organizations should design IoT projects with knowledge management in mind. If they don’t, they will spend too much time searching for outcomes and key performance indicators that become easy to identify when making the connection between knowledge, the need for information and IoT as a validation for the first, and an input to the second. KM provides a framework for IoT implementation related to collecting data, sense-making and tying data to business outcomes.

More from Serious Insights on IoT.

Daniel W. Rasmus

Daniel W. Rasmus, Founder and Principal Analyst of Serious Insights, is an internationally recognized speaker on the future of work and education. He is the author of several books, including Listening to the Future and Management by Design.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.