NXP Semiconductors emerged as the winner in the Best AI and Edge Computing Solution category at the 2025 Sensors Converge Conference and Expo in Santa Clara, CA, on June 26.
NXP Semiconductors won with its eIQ Time Series Studio, an end-to-end AI development tool that is designed to democratize time series machine learning for edge devices. Winners were selected from an outstanding number of submissions and evaluated by a panel of industry judges. The annual Best of Sensors awards are sponsored by Questex's Sensors Converge and Fierce Electronics.
The eIQ Time Series Studio addresses a complex challenge in machine learning: developing effective time series models for resource-constrained edge devices. Time series data consists of sequences recorded at consistent intervals and is fundamental to applications from predictive maintenance to anomaly detection, yet it traditionally requires extensive data science expertise.
At the heart of eIQ Time Series Studio lies its automated machine learning (autoML) capability, which transforms what once required weeks or months of manual development into a process that can be completed in hours. The platform automates critical tasks, including signal processing pipeline selection, model architecture determination, hyperparameter tuning, and feature engineering. When presenting the award, judges noted that “its offline-first, no-cost model is a major differentiator in a landscape crowded with cloud-first tools,” emphasizing the tool’s ability to enable non-experts to deploy robust sensor-based models rapidly.
The studio supports an extensive range of sensor inputs, including temperature, vibration, pressure, microphone, time of flight, voltage, and current sensors, as well as multimodal sensor fusion. This allows developers to tackle applications across various environments, such as industrial, automotive, smart city, and smart home, ranging from arc fault detection to battery health prediction.
One innovative feature is its integrated data intelligence utility, which evaluates dataset balance, identifies redundant channels, and recommends optimal sampling frequencies and window sizes. This addresses common challenges, such as over-sampling or unbalanced datasets that compromise model effectiveness. The utility can recommend reducing sampling rates to 1/16 of the original while maintaining accuracy and reducing computational requirements.
The development workflow within eIQ Time Series Studio is designed for accessibility. Developers can import custom time series data with multiple viewing options, including raw, temporal, statistical, and spectral perspectives. The automated training process generates multiple model options that can be sorted by accuracy or memory footprint, allowing developers to select the optimal balance for their specific hardware constraints. The platform then provides emulation capabilities to test models in a virtual edge environment before deployment.
What sets eIQ Time Series Studio apart is its offline-first approach, prioritizing data privacy by eliminating the need to upload proprietary data to cloud platforms for training. This design choice, combined with the tool being completely free with no recurring subscription fees, removes significant barriers to adoption for organizations of all sizes. The platform generates deployment-ready libraries compatible with MCUXpresso and Code Warrior IDEs, requiring only two API calls for implementation, one for initialization and another for inference.
The tool’s impact extends beyond individual projects to broader market transformation. By reducing the barrier to entry for time series AI development, eIQ Time Series Studio enables organizations to leverage intelligent edge devices that are faster, more secure, and more energy- and cost-efficient than cloud-dependent alternatives.
NXP Semiconductors is a global semiconductor company that enables secure connections and infrastructure for a smarter world. With the eIQ Time Series Studio, introduced in October 2024 and available as part of the free eIQ toolkit, NXP continues to advance edge AI capabilities, making sophisticated machine learning accessible to developers regardless of their data science background.
To discover other innovative sensor technologies and see the complete list of 2025 Best of Sensors Award winners, visit the Best of Sensors Awards announcement.