AI is a broad space with a spectrum of techniques to support a diverse set of use cases, each requiring different forms of intelligence and computational strategies. Chain of thought conversations and content generation are powered by GenAI. Machine learning models forecast equipment failures and energy demand using historical and real-time data, while optimization and simulation support efficient planning and decision-making. Autonomous systems use reinforcement learning and intelligent automation. Supervised learning supports segmentation and classification, while recommendation engines blend collaborative filtering with predictive analytics. For perception, AI interprets sensor data to understand the environment. Anomaly detection relies on statistical and unsupervised methods to identify deviations. Finally, knowledge graphs and semantic AI uncover insights across complex systems, together forming the intelligent backbone of digital twin platforms. 

Willow strategically applies the full spectrum of AI methodologies across features, selecting the most effective technique for each use case. 

Types of AI chart

AI Design Patterns in Willow 

Willow employs a rich set of AI design patterns, from generative models and supervised learning to forecasting and optimization. Each is tailored to specific use cases, from content generation, asset classification, to predictive maintenance. Let’s walk through examples of these AI methodologies getting applied in Willow. 

  1. Prediction and forecasting: Willow’s Activate Technology uses leading indicators to anticipate high-cost events such as equipment failures or food waste. Vibration analysis and leak detection powered by IoT signals from third-party integrations enables proactive maintenance and replacement scheduling. These predictive capabilities help facility managers avoid costly downtime and extend asset lifecycles. 
  2. Planning: Willow Skills provide out-of-the-box and customizable logic for managing HVAC systems, refrigeration units, water systems, and more. Users can tailor and extend these Skills using Activate Studio, allowing for flexible planning and control strategies that adapt to the unique needs of each building or campus. 
  3. Decision intelligence: Willow Insights deliver diagnostics and impact scoring with KPIs like Avoidable Cost, helping users prioritize actions based on potential outcomes. For example, Willow supports dynamic load shedding and battery energy storage system (BESS) charge/discharge optimization, enabling smarter energy management decisions that balance cost, sustainability, and operational needs. 
  4. Autonomous systems: Active Control closes the loop between detection and action. When an anomaly is identified, such as a temperature spike or water leak, the system can automatically adjust a setpoint or close a valve. This reduces the need for manual intervention and ensures rapid response to critical events. 
  5. Segmentation and Classification: Willow uses Machine Learning and computer vision to classify devices and populate its Knowledge Graph. This graph is built on a rich ontology using DTDL (Digital Twin Definition Language), modeling systems from HVAC and plumbing to Passenger Boarding Bridges for airplanes. The ontology is extensible across domains, from office buildings to hospitals and laboratories, where specialized assets like fume hoods can be accurately modeled and managed.   
  6. Recommendation Systems: Willow Dashboards surface relevant insights at login, such as occupancy trends or sustainability metrics. Continuous commissioning skills run diagnostics and generate actionable recommendations, guiding users toward corrective measures for failing systems or inefficiencies.   
  7. Perception: AI vision systems and device signals ingested from third party integrations are processed to infer presence and foot traffic. This data feeds into occupancy dashboards that help building operators understand usage patterns, optimize space utilization, and improve safety. Device MAC address tracking as a proxy for user presence and sensor fusion further enrich these insights, offering a multi-layered view of space usage. 
  8. Intelligent Automation: Intelligent sub-graph generation with GenAI facilitates fast onboarding with traditional Single Line Diagrams and Test and Balance reports. Another example is asset visualization in 2D/3D powered by the knowledge graph that correlates capabilities and relationships across systems. This allows for automated workflows, contextual awareness and streamlined operations. 
  9. Anomaly Detection and Monitoring: Willow continuously analyzes telemetry streams to identify deviations in energy efficiency, data quality, and equipment behavior. Insights and KPIs help quantify the severity of anomalies, enabling faster triage and resolution and maintaining system health. 
  10. Knowledge Discovery with Conversational Interfaces: Willow Copilot is a GenAI-powered natural language interface grounded in customer data. Users can upload documents and images to generate summaries for reporting, troubleshooting, or inspections. Multilingual and semantic search support enables intuitive interactions for tasks like equipment troubleshooting and Retrieval Augmented Generation (RAG) with documents. Data in the Knowledge Graph enables generative responses, helping users uncover insights, understand system behavior, and make confident decisions.

 Closing thoughts 

Willow’s use of AI spans a comprehensive list of techniques that range from traditional machine learning and rule-based logic to cutting-edge generative models and autonomous control systems. Each technique is carefully applied to specific use cases, delivering tangible value in the form of predictive insights, operational efficiency, and intelligent automation. By weaving these capabilities into a unified platform, Willow empowers organizations to manage built environments with unprecedented intelligence and agility. 

Read on and explore more of Willow’s AI features below:

From Blueprints to Bytes: How Gen-AI Subgraph Generation is Transforming Building Digitization

How Willow Copilot Tackles the Field Knowledge Gap

Capital Planning with Willow Copilot – A Day in the life of a CIO

Activate Studio: the ‘What’ and ‘Why’