Ekkono makes connected products smart by embedding advanced machine learning into connected devices. Ekkono delivers a software library, specially designed for use in products at the edge of the network. By applying Ekkono’s SDK (Software Development Kit) to products with sensors, customers can develop smart, predictable, self-learning functions for their connected products.
Ekkono was founded in 2016, based on seven years of research at the University of Borås. It is a software company that makes machine learning for IoT – also known as Edge Machine Learning. Ekkono helps companies to create new values and services through their connected products.
Ekkono is unique in that its solution can be used on very small devices such as microprocessors and microcontrollers. Another strength is its ability to use incremental learning, whereby the machine learning module is continually trained on the device. This means that the solution not only runs static, generic models based on general cloud data, but learns progressively so that individual models are created for each unit during use.
Typical applications for Ekkono’s solutions are virtual sensors, status monitoring, device modelling, predictive maintenance, control and alarming.