The chat responses are generated using Generative AI technology for intuitive search and may not be entirely accurate. They are not intended as professional advice. For full details, including our use rights, privacy practices and potential export control restrictions, please refer to our Generative AI Service Terms of Use and Generative AI Service Privacy Information. As this is a test version, please let us know if something irritating comes up. Like you get recommended a chocolate fudge ice cream instead of an energy managing application. If that occurs, please use the feedback button in our contact form!

Provide Feedback

Feedback sent

Thank you for your feedback!

Provide Feedback

Please log in to provide feedback
Skip to content
Siemens Xcelerator Marketplace

Terminology

Senseye is a predictive maintenance system designed to monitor industrial machinery to raise insights on the condition, and any potential degradation, of these machines. This page is a collection of common terms used within Senseye.

Asset

An object being monitored by Senseye, in most cases this is an individual piece of industrial machinery such as a motor or gearbox. However, if limited monitoring data is available this can be a collection of machines, such as a drive chain, or if an abundance of monitoring data is available can be components of an industrial machine, such as motor windings.

Assets form the notification level of Senseye, these notifications are known as cases, and are raised when an asset's Attention Index is raised to the medium or high level. Only one case can be open against the asset at a time. Cases contain evidence in the form of insights generated from the measures associated with the sensors attached to the asset.

Assets are organized in the hierarchy, where they can be children of sublevels. An asset cannot be the child of another asset.

Assets are identified by a unique UUID.

Attention Index

A score for the amount of attention needed on an asset. This score is generated using a machine learning algorithm based on insights raised by Senseye. When the Attention Index reaches Medium or High a case is raised on the asset.

The Attention Index is always between 0 and 1 inclusive, and categorized below:

  • Normal: 0 - 0.25
  • Low: 0.25 - 0.5
  • Medium: 0.5 - 0.75
  • High: 0.75 - 1

Case

A notification within Senseye. Cases are opened when the Attention Index on an asset goes into Medium or High. Each case has associated case evidence, containing the insights that contributed to the case being opened. Only one case can be open on an asset at a time.

Once a case has been handled it can be closed. Closing a case requires feedback to be provided using the latest feedback questions.

Cases can be associated to a user via case acknowledgement.

Cases are identified by a unique UUID.

Event

Additional timestamped information against an asset. Events are used to record timestamped information such as investigation notes, logs of maintenance work, or recording of KPIs. There are multiple types of event, the most commonly used being:

  • Work Event: Record of Maintenance Work Done
  • Note: Additional free type information
  • Manual Measurement: Log of an manual measurement
  • KPI: Record of resources saved/used

Events are identified by a unique UUID.

External ID

A unique mapping between an object in Senseye and a unique reference. Most commonly used to map an UUID, such as an asset ID, to a client's ID for the object, such as the machine's location ID.

Each external ID has three parts:

  • Type: what system the mapping relates to
  • Internal ID: a unique internal reference, commonly a UUID
  • External ID: a unique external reference string

Hierarchy

Structure of an organization. Each hierarchy is made up of nodes, of which there are four types:

  • Organization
    • Representative of the user's tenant
    • Always at the top level of the hierarchy
  • Sublevel
    • A level of the hierarchy
    • Used to add structure/grouping to the hierarchy, similar to a folder
  • Asset
    • The notification level of the hierarchy
    • Can be a child of a sublevel or an organization
  • Sensor
    • The data level of the hierarchy
    • Can only be a child of an asset

All nodes can be identified by a unique UUID.

For sublevels, assets, and sensors external IDs of type NODE can be setup. This can be used to query the Senseye hierarchy based on client reference IDs rather than UUIDs.

Insight

Detection of a change on a measure, also called patterns. Insights are the primary output of Senseye's algorithms and are detections of changes on measures. There are several different types of insights, corresponding to different forms of detection:

  • Anomalies
  • Trends
  • Threshold Violations
  • Forecasted Threshold Violations
  • Degradations
  • Failure Matches
  • Missing Data (Data Issue)
  • Flatlining (Data Issue)

Missing Data and Flatlining insights cannot raise a case but are used for detection of data issues, rather than machine degradation.

Insights are identified by a unique UUID.

Measure

An individually named time series. Measures form the basis upon which insights are detected. They are automatically generated from the time series data sent to a sensor, appearing shortly after first being sent to the sensor.

Measures are identified with their fully qualified name (FQN), which takes the form of measure:{sensor-id}:{measure-source-name}. The measure source name is the name of the measure as sent in the time series data or is proscribed in the case of measures extracted from frequency data. See machine data for more information.

Organization

A customers tenant. Senseye is multi-tenanted, with each of the individual tenants being referred to as an organization. Each organization is a self contained space, with it's own hierarchy, cases, and users.

An organization is identified with a group UUID, where group| prefixes the organization's UUID. In all cases an organization is associated to the user's authentication and the organization will not need to be specified.

Sensor

A logical grouping of measures. Sensors were originally meant to reflect physical sensors on assets which may report multiple measures. Today sensors can be any logical grouping of measures that need to be associated with an asset.

Sensors have an 'ingest' property which determines what type of data they can accept. A 'timeseries' sensor can ingest raw time series data, for example scalar data measured directly from temperature, oil, pressure sensors. Alternatively, a 'frequency' sensor will only ingest frequency spectra files, which will undergo postprocessing to extract measures.

Sensors are attached to assets to join the hierarchy, once attached the data from the measures on the sensor are used for monitoring the asset.

Sensors that are not attached are known as unattached sensors. They can be found by querying for the unattached sensors in the organization.

Each sensor can only be attached to one asset.

Measures are automatically generated based on the machine data sent to the sensor.

Sensors are identified with a unique UUID. For the purposes of sending machine data an external ID of type SENSOR, otherwise known as a mapping ID, can be setup.

Sublevel

A level of the hierarchy under which further sublevels or assets can be contained. Sublevels are often used to organize assets into logical groupings within Senseye, and often represent physical locations, such as a site or area.

User

An individual's account in Senseye. Users action cases and feedback to Senseye. Where a user takes an action relating to a case or leaves an event on an asset they will be recorded as the author.

Users are identified by a unique ID, constructed from identity provider specific details joined by |.

UUID

A 128-bit identifier as specified in RFC 4122.

UUIDs are specified in the 8-4-4-4-12 format, like xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx.

Contact us