Concept: San Jose’s technology startup Edge Impulse has leveraged TinyML and AutoML technologies to bring AI capabilities to microcontrollers. TinyML is an ML technique that can be implemented in low-energy systems including sensors to perform automated tasks. It optimizes ML models for embedded devices that are resource-constrained. AutoML focuses on data acquisition and prediction aspects of ML. Edge Impulse has also developed a cloud-based platform to enable developers to build AI models targeting microcontrollers.

Nature of Disruption: TinyML inspires the use of about 99% of sensor data that is discarded presently due to cost, bandwidth, and power constraints. It enables the processing of various types of input data including radar, motion, electromagnetic fields, audio, images, and device logs. Edge Impulse enables developers to collect or upload training data from devices, label the data, train a model, and launch and monitor models in a production environment. The new platform provides software development kits (SDKs), clients, and generated code to the developers as open-source under an Apache 2.0 license. It supports the development of ML for sensors, audio, and computer vision, specializing in TinyML industrial applications including predictive maintenance, asset tracking, and monitoring, and sensing. One of the key advantages of running AI in embedded systems is performing predictions without depending on the cloud or external services. The models become a part of the firmware flashed into the devices.  The startup mentions that the new platform can run on both Windows and MacOS platforms. It claims to leverage a compiler that can compile models to C++ to increase the efficiency of models trained on its platform, which can reduce RAM usage by 25% to 55% and storage usage by up to 35%.

Outlook: A microcontroller is a small and low-cost microcomputer designed to perform the specific tasks of embedded systems. It consists of the processor, the memory, serial ports, and peripherals. Developing AI functionalities for microcontrollers is a complex process. Edge Impulse’s platform enables the implementation of ML models on microcontroller embedded devices. TinyML technology has applications in various industries including manufacturing, retail, and agriculture. In the manufacturing sector, TinyML can help prevent downtime by alerting users to perform preventative maintenance based on equipment conditions. In December 2021, the startup raised $34M in a Series B funding led by Coatue with participation from several investors including Canaan Partners, and Acrew Capital. It aims to use the funding to expand its employee size and hardware partner network.

This article was originally published in Verdict.co.uk