Smarter, Safer, and Efficient Edge Data Management

Overview

ITTIA DB Lite provides microcontrollers (MCUs) with a structured and efficient way to manage data directly on the device, transforming them from simple control units into intelligent data processors. Instead of relying solely on external systems, ITTIA DB Lite enables MCUs to collect, organize, and store sensor readings in real time, while supporting queries, filtering, and aggregation with deterministic performance under tight memory and power constraints. Features such as time-series storage, data compression, and secure access allow large volumes of sensor data to be handled efficiently, while reducing flash wear and conserving bandwidth by transmitting only meaningful insights rather than raw streams. This not only improves reliability and responsiveness in edge applications but also enables on-device AI workloads like anomaly detection and predictive maintenance, making ITTIA DB Lite essential for modern data-centric MCU applications in industrial automation, automotive, healthcare, and smart energy systems.

Case study

This case study demonstrates how the STM32U5 series microcontrollers integrated with ITTIA DB Lite enable secure, low-power, and intelligent data management for portable medical devices and smart energy applications. STM32U5 delivers a Cortex-M33 core up to 160 MHz, up to 2 MB Flash and 786 KB SRAM, alongside advanced low-power modes and Arm TrustZone-M security. ITTIA DB Lite adds a lightweight relational database with querying abilities, time-series storage, compression, and AI model integration. Together, they transform MCUs into secure edge data nodes capable of real-time analytics and AI workloads under tight energy and memory budgets.

Use Case: Wearable Health Monitoring Device

Challenge

A medical technology company needs to design a wearable cardiac health monitor that:

• Collects ECG and accelerometer data 24/7.
• Runs in low-power mode to maximize battery life.
• Detects anomalies locally (arrhythmia, irregular heartbeat).
• Stores several days of sensor history securely on-device.
• Shares critical events securely with other part of the embedded system.

System Architecture

• MCU: STM32U585 (Cortex-M33, TrustZone-M, AES/PKA cryptography)
• Sensors: ECG sensor (SPI), 3-axis accelerometer (I²C)
• Comms: BLE 5.2
• Software Stack:
  • OS: FreeRTOS/ Eclipse ThreadX RTOS /Zephyr
  • Database: ITTIA DB Lite
  • AI Model: NanoEdge AI Studio, LiteRT (TensorFlow Lite Micro) anomaly detection model
  • Visualization: ITTIA Analitica
  • Data distribution to MPUs: ITTIA Data Connect
  • Security: STM32U5 TrustZone + ITTIA DB Lite secure storage 

Implementation

1. Schema Design

   CREATE TABLE ecg_data (
       sample_time TIMESTAMP PRIMARY KEY,
       lead1       FLOAT NOT NULL,
       lead2       FLOAT NOT NULL,
       accel_x     FLOAT NOT NULL,
       accel_y     FLOAT NOT NULL,
       accel_z     FLOAT NOT NULL
   );
   
   CREATE TABLE events ( 
       event_time TIMESTAMP PRIMARY KEY,
       score      FLOAT NOT NULL,
       severity   VARCHAR(32) NOT NULL
   );

    

2. Data Logging (C API Example) 

   #include "ecg_database.h" // generated from schema
   
   int main() {
      db_init_ex(DB_API_VER, NULL);
      open_ecg_database("ecg", NULL);
      // ...
   }
   
   void log_ecg_data(uint64_t ts_ms, float l1, float l2, float ax, float ay, float az) { 
       db_t db;
       db_connect(&db, "ecg", "ingestion", NULL, NULL);
       const ecg_data_row_t row = {
           .sample_time   = ts_ms * 1000, // to microseconds
           .lead1         = l1,
           .lead2         = l2,
           .accel_x       = ax,
           .accel_y       = ay,
           .accel_z       = az
       };
       put_ecg_data(db, row);
       db_disconnect(db);
   }

   • Data written in batch mode (every 1 second) to conserve flash cycles.
   • Compression reduces storage demand by ~45%. 

    

3. Local AI Inference

   Every 5 seconds, an MCU callback function:

   • Receives the last 500 ECG samples.
   • Data is normalized and fed to an AI arrhythmia detection model.
   • Results are unconditionally captured in an events table.

    

4. Secure Data Sharing

   • When connected via BLE, only anomaly events are transmitted.
   • ITTIA DB Lite uses STM32U5 hardware crypto engine to encrypt storage and enforce secure access under TrustZone-M.
   • Prevents unauthorized access to sensitive patient data.

    

Results

Benefits

1. Ultra-low power: STM32U5’s tickless RTOS + ITTIA DB Lite batch writes extend battery life.
2. Security by design: TrustZone-M + encrypted storage protect patient data.
3. Local intelligence: AI inference reduces dependence on cloud/mobile.
4. Bandwidth efficiency: Only anomalies/events transmitted, not raw streams.
5. Scalability: Same schema scales to STM32H7 for high-performance devices.

Conclusion

By leveraging the energy-efficient STM32U5 platform with ITTIA DB Lite, developers can build secure, intelligent, low-power wearable and portable devices. This architecture enables continuous sensor monitoring, local data management and AI processing, as well as secure event reporting, ideal for medical devices, smart energy meters, portable diagnostics, and other battery-powered IoT systems where both data privacy and power efficiency are paramount.