We are pleased to announce a significant upgrade to the Favoriot Platform. Favoriot is no longer positioned solely as an IoT platform. With the introduction of native Machine Learning capabilities, Favoriot now operates as a complete AIoT Platform, enabling predictive, adaptive, and data-driven intelligence directly on top of IoT data. This feature is called “Favoriot Intelligence“.
This upgrade reflects where the industry is heading and prepares users for what comes next. IoT data is no longer just for monitoring and visualisation. It is meant to learn, anticipate, and support better decisions over time.
From Connected Data to Intelligent Systems
The new Machine Learning (ML) Engine allows IoT data collected within Favoriot to be analysed and learned automatically, without the need to build or manage external ML pipelines.
Machine learning models are created, trained, scheduled, and applied entirely within the Favoriot platform. Users configure models through a guided interface by selecting data sources, algorithms, and parameters. No complex ML coding is required.
This feature is available exclusively for Developer accounts.
What the Machine Learning Engine Enables
The ML Engine is tightly integrated with Favoriot’s data ingestion, storage, dashboards, analytics, rule engine, and workflow tools. This allows machine learning results to move seamlessly from training to real-time usage.
Key Objectives
- Predict future trends and system behaviour using historical IoT data
- Reduce dependency on static, threshold-based rules
- Improve accuracy through data-driven learning
- Deploy models consistently across devices, projects, and data streams
Supported Machine Learning Categories
1. Time Series Forecasting
Predicts future values based on historical, time-ordered data.
Algorithms
- Linear Regression
- ARIMA
- LSTM
Common Use Cases
- Energy consumption forecasting
- Environmental sensor prediction
- Traffic or system load estimation
2. Unsupervised Learning
Identifies patterns, clusters, or anomalies in unlabeled data.
Algorithms
- K-Means
- DBSCAN
- Gaussian Mixture Model (GMM)
Common Use Cases
- Device behaviour clustering
- Anomaly detection
- Early indicators for predictive maintenance
3. Supervised Classification
Classifies IoT data into predefined categories using labelled data.
Algorithms
- Support Vector Machine (SVM)
- Logistic Regression
- Random Forest
- XGBoost
Common Use Cases
- Device health classification
- Fault categorization
- Security or intrusion detection
4. Supervised Regression
Predicts continuous numerical values.
Algorithms
- SVM
- Linear Regression
- Random Forest
- XGBoost
Common Use Cases
- Remaining Useful Life (RUL) prediction
- Power or resource usage estimation
- Sensor value prediction
System-Level Benefits
- Simple Configuration
Models are set up through guided forms without complex ML development. - Predictive Intelligence
Detect issues and trends earlier, rather than reacting after failures occur. - Adaptive Learning
Models evolve as new data becomes available. - Scalable Deployment
Apply the same model logic across multiple devices and projects. - Fully Integrated
ML outputs can be visualised in dashboards, trigger alerts, or drive rule-based workflows.
Model Creation Workflow (Overview)
Creating a machine learning model follows a structured and clear flow:
- Identify the Model
Define the model name and purpose. - Configure Data Source
Select category, historical date range, and timezone. - Model Configuration
Choose algorithms and parameters based on the ML category. - Training
Start training immediately or schedule it to run at defined intervals. - Evaluation
Review graphs, performance metrics, and processing details. - Deployment
Use results in dashboards, rule engines, or download models for edge or gateway inference.
Scheduling and Continuous Learning
Models can be automatically retrained to stay up to date with incoming data.
Supported schedules:
- Interval-based (up to every 12 hours)
- Daily
- Weekly
- Monthly
This ensures models remain relevant as data patterns change.
Visualising and Using ML Results
Machine learning outputs can be:
- Displayed directly in analytic dashboards
- Applied in Favoriot’s rule engine workflows
- Used for real-time or offline inference
Developers can download trained models and deploy them on gateways or devices using Python 3.12 or later, following a structured inference pipeline to preserve accuracy.
A Strategic Step Forward
This upgrade marks a meaningful shift in how Favoriot supports modern IoT solutions. By embedding machine learning directly into the platform, developers can move beyond monitoring toward systems that learn, anticipate, and act with context.
Favoriot now supports the whole journey:
From data collection to learning to prediction to action.
We invite Developer account users to explore the new Machine Learning features and begin building intelligent IoT systems that are ready for both today’s needs and tomorrow’s expectations.
For detailed setup guidance, please refer to the Machine Learning walkthrough videos and documentation available within the platform.
FAVORIOT Intelligence References
- Favoriot Intelligence (Machine Learning) Documentation
- Official Announcement -Favoriot Evolves into an AIoT Platform with Built-in Machine Learning
- Why Favoriot Chooses Machine Learning Over LLMs
- Machine Learning vs Deep Learning in IoT
- Traditional Machine Learning vs Large Language Models (LLMs)
- Why Favoriot’s Built-in Machine Learning Matters for AI Researchers and IoT Developers
- Why Favoriot’s ML Infrastructure Reduces Costs
- AI and ML in IoT, Explained Without the Jargon
- Start Anomaly Detection in IoT with Less Data
- AI and ML in IoT, Explained Without the Jargon
- The Key Differences: Favoriot’s Rule Engine 2.0 and AI Agents
- The Role of Machine Learning in IoT Systems
- Favoriot’s Rule Engine 2.0: A Structured Approach to IoT Automation
- A Quiet Shift Is Coming to the Favoriot Platform
Leave a Reply