SensiML toolkit helps develop IoT endpoints using NXP MCUs
March 11, 2020
Developers of IoT endpoint using NXP MCUs can access AI development tools for rapidly creating efficient and robust time-series sensor algorithms by using an analytics toolkit from Oregon-based SensiML.
The developer of AI tools for building intelligent IoT endpoints has announced support for Dutch electronic giant NXP's i.MX RT portfolio of crossover microcontrollers via its analytics toolkit and plans to support out-of-the-box data collection with the NXP i.MX RT1050 evaluation kit and additional i.MX RT-based development boards in the future.
SensiML delivers a complementary end-to-end AI development tool that includes data collection, labelling, feature extraction, ML (machine learning) classification and embedded auto code generation. Through the partnership with NXP, products from the combined companies give IoT endpoint developers a complete method for AI-based sensor algorithm creation and deployment.
Because so much of the AI work is automated by the toolkit, these algorithms can be developed up to five times faster than those using traditional hand coding. This allows specialised IoT products to be successfully developed without the need to invest in large teams of data scientists and firmware engineers.
"Developing accurate industrial IoT sensor algorithms supporting anomaly detection and predictive maintenance is extremely complex considering the dynamic nature and many sources of variability with which to contend," said Markus Levy, director of AI and machine learning technologies at NXP. "Tools such as the SensiML analytics toolkit help by alleviating the developer from having to understand the data science behind ML algorithms and subsequent coding and optimisation needed to implement in efficient IoT firmware. This translates into significant efficiency gains during development and allows developers to focus on collection of labelled datasets specific to their application."
The i.MX portfolio of crossover MCUs and the i.MX RT1050 evaluation kit (pictured) are available now from NXP, and the analytics toolkit is available now from SensiML. Support for library code generation on i.MX RT crossover MCUs is available today. Native support for data collection on the i.MX RT1050 evaluation kit will be offered later in the first half of 2020.
"With its powerful 600MHz Arm Cortex-M7 processor and many interface options, the i.MX RT1050 MCU fills a key segment for SensiML customers seeking performance MCUs for IoT endpoint applications like predictive maintenance, process control and structural health monitoring," said Chris Rogers, CEO at SensiML. "Our automated AI code-gen toolkit enables developers using the i.MX RT1050 to rapidly and easily build complex multi-sensor recognition algorithms for such advanced applications."
SensiML, a subsidiary of QuickLogic, offers software that enables low-power IoT endpoints that implement AI to transform raw sensor data into meaningful insight at the device itself. The analytics toolkit provides an end-to-end development platform spanning data collection, labelling, algorithm and firmware auto generation, and testing. The toolkit supports Arm Cortex-M class and higher microcontroller cores, Intel x86 instruction set processors, and heterogeneous core QuickLogic SoCs and QuickAI platforms with FPGA optimisations.