Architecture overview

The following diagram and table describe the architecture overview for IBM® TRIRIGA® Building Insights and the integrations with IBM® TRIRIGA®.

The Building Insights architecture consists of following components, which includes components from IBM TRIRIGA.

Component group | Component   |Description   
--------------- | ----------- | ------------ |

TRIRIGA | TRIRIGA Data | TRIRIGA is the integrated workplace management system (IWMS) that is supported by Building Insights. TRIRIGA is the system of records for all portfolio data such as organization hierarchy, geo hierarchy, and user-related information. TRIRIGA is also the system of record for all space-related data that includes details of buildings, floors, workspaces, and details related to workspace allocation. TRIRIGA | Servlet proxy | Acts as proxy for all requests that go from TRIRIGA to Building Insights. Called by UI requests, for data to render on the UI. Connects to Building Insights running on cloud, fetches the data and returns the data back to the UI. This component is deployed on TRIRIGA through Object Management Package.| TRIRIGA | OSLC API | Exposes TRIRIGA data through REST API. The exposed data includes building, space, allocation, and portfolio data. This component is deployed on TRIRIGA through the Object Management Package.| Building Insights UI | Building Insights Perceptive app | Built by using Watson IoT components, the Perceptive app is deployed on TRIRIGA by importing the Object Migration Package. Provides space-related insights at building level, floor level, and organization level to a Space Planner. The Space Planner can perform time-based filtering and can get the details of different KPIs, such as occupancy rate, on the UI dashboard. The UI also provides navigation function to navigate across floors and buildings. | Integration | Sensor onboarding API | Building Insights integrates with IoT Sensors, and provides APIs to onboard the sensors and sensor gateways, and to associate them to a space in TRIRIGA. Also, the APIs create the related devices on Watson IoT Platform to enable device to sensor (and gateway) connectivity and data transfer. | Integration | TRIRIGA integration API | Periodically pulls space-related data from TRIRIGA and stores the data in a knowledge graph. Provides API for servlet proxy to call, to provide the required data to the UI at floor level and building level. | Integration | Cisco DNA Spaces data receiver, processor, xy transformation | Building Insights integrates with Cisco DNA Spaces, which uses WIFI data to provide the occupancy data at a floor level. The data is received from the DNA Spaces Firehose API, stored and processed to understand the organization that relates to the person present on the floor. This data activity frequency is every 15 minutes. | Integration | Knowledge graph | Stores all the metadata about buildings, floors, spaces, which Building Insights requires. Provides the flexibility to add any available metadata (for example, sensor that is attached to a space).| Integration | IoT Sensor Data Processing | Processes data from desk level sensors and maps the occupancy count to business units every 15 minutes. | Integration | Sensor data processing | Is a custom component that is needed for some sensors (for example, motion data and CO2), to convert raw data from sensors to occupancy count.| Integration | Location mapper | Maps the Cisco DNA floor maps to the TRIRIGA floor map and transforms the XY coordinates of a floor, with the reference offset. Handles input of the reference offsets and units of measure during the synch building process. | Integration | Webhook API | Building Insights provides a mechanism via Webhook API to obtain the user ID mapping. The Webhook API returns a JSON response to the Building Insights component when invoked. | Watson IoT Platform | User directory | Stores Building Insights user information, including the Administrator user. Uses AppID that is provided by Watson IoT Platform. | Watson IoT Platform | File storage | To store files, Building Insights uses Watson IoT Platform Cloud Object Storage (COS). Includes log files, AI reports, and some of the configuration data. | Watson IoT Platform | Data lake | Building Insights uses PostgreSQL data lake that is provided by the platform. The data is stored in this relational database for all the time series data. The data includes raw data that is received and data output of calculations. | Watson IoT Platform | Connect and collect | Core component of Watson IoT Platform. The time series data collector for all the data, which includes data pulled from TRIRIGA. Also, data that is received from DNA Spaces, IoT Sensors, and data that is processed in the AI components. | AI Framework and models | AI Framework | Contains sub components that are required for scheduling AI jobs to call the AI model and to do pre-aggregations and KPI calculations such as occupancy rate calculations. | AI Framework and models | Pre-aggregations | Data that is received from TRIRIGA and DNA spaces are pre-aggregated at daily granularity at organization, floor, and building level. | AI Framework and models | KPI Calculation | Calculates KPI's like occupancy rate and calculates the output of AI models like weekend detection and anomaly detection. Calculations are based on pre-aggregated data. KPI calculation is performed for each organization in each floor and considers the time zone of that building. | AI Framework and models | AI and Analytical Models | The following AI models are deployed. Anomaly Detection Model detects anomalies on occupancy count. Weekend Detection Model automatically determines the weekly holiday for an organization. Workpoint Availability Pattern Recognition Model calculates normal working hours for workpoints that is used for workpoint frequency rate calculations. Data Quality Analysis Model provides detailed data quality report on the data that is used by some models. For more information about the AI models, see AI and Analytical Models|