Founded by Zach Shelby and Jan Jongboom, Edge Impulse is the go-to platform for embedded machine learning (TinyML). Edge Impulse is well positioned to become a leader in edge intelligence, leveraging the convergence of low-cost, high-performance microprocessors and low-power, wide area networks such as LoRaWAN®, to create intelligent, extremely low-power and wireless devices. Edge Impulse is on a mission to enable developers to create the next generation of intelligent devices using embedded machine learning in industrial, enterprise and human centric applications.
Today 99% of data generated by sensors is discarded due to cost, bandwidth or power constraints. Computers have become so ubiquitous at the edge, device sensors can read anything you might want to know about your business operation or home. They’ve also become powerful enough that embedded firmware processes like machine learning require very little power from the device’s battery. However, sending that data back up to the cloud or to an on-premise server can cost both money and battery life for the device. In order to use cheap low-power networks like LoRaWAN® the device must only transmit tiny amounts of data (up to 50 bytes per transmission).
Edge Impulse provides a platform that allows all developers the ability to create embedded machine learning algorithms for their devices. On the platform developers can manage data sets for their embedded computers as well as design, train, test and deploy models for those computers. Of the vast amounts of data exhaust generated by today’s embedded computing devices, machine learning models can infer relevant or actionable data. The device can then take action automatically the way a thermostat adjusts temperature. Alternatively, just that relevant data can be sent back to the cloud, such as sending a message to a factory manager that an expensive piece of machinery is likely to fail. Data-heavy sensors like audio, radar and images can be digested and turned into simple messages, saving energy, money, and importantly in many circumstances, time.