From a high-speed airplane flying at a speed of 550 mph to an autonomous car doing a modest ten mph, sensor-based IoT services hold great promises of making "everything smarter". This is good reason for any organization to delve into the innovative networking and interconnected paradigm created by IoT. Much of the sensor based IoT initiatives are geared towards low-cost and energy-efficient hardware for systems. However, finding the right solutions to manage and utilize the massive volumes of data produced by these systems has snowballed into a challenge for the CIOs. “This requires the individual elements to get smarter, which can be done using machine learning. Smarter things is a necessity, or the great prospect of IoT will be lost in a tsunami of sensor data and business cases that are broken by the labor cost for manual supervision of this data flood,” says Jon Linden, the CEO of Ekkono Solutions. By applying machine learning on the sensor data—data that cannot be sent to the cloud due to its sheer volume—in real time, Ekkono Solutions makes the devices smarter through self-learning, anomaly detection, predictions of future faults, and by identifying what makes them work optimally and be more intuitive. Based on seven years of machine-learning research at the University of Borås, the firm has developed and delivers a lightweight embedded software solution for running advanced machine learning at the edge.
While its unique design makes it resource efficient with an unparalleled small footprint which enables it to run at the edge, its all - software solution design empowers organizations to be not dependent on hardware dependencies or requirements
Curated and relevant data is fed by Ekkono’s solution to the cloud such that IoT deployments do not have to rely upon averages of historical data uploaded to the big data haystack. The uploaded data can be leveraged by organizations for further analysis and cross-referencing across the installed base. Further, the firm devises methodologies that can learn the behavior of a specific device and can then provide customized results accordingly. Ekkono’s solution is applicable both for industrial and consumer IoT. While the firm delivers its solutions to machine manufacturers and automotive companies, it also provides a better user experience on the consumer IoT side by facilitating intuitive and self-learning things.
Devised as an easy-to-use embedded advanced analytics engine for IoT, Ekkono’s design is platform-independent and runs in virtually any environment. “While its unique design makes it resource efficient with an unparalleled small footprint which enables it to run at the edge, its all-software solution design empowers organizations to be independent of hardware or other requirements,” mentions Linden. Ekkono’s product, which is designed as an SDK (software development kit) with embedded code, manages the entire workflow from data pipelining and model training, to model optimization, execution and re-training. The firm works with system integrators to implement its technology for the customers. Being the pioneers, the firm gets involved in the data and business understanding phase and running a pilot project to understand its customers' needs to streamline and validate the data hypothesis. "We recommend our customers not to boil the ocean at once, but solve one problem after another to really get pay-back on connecting their products,” mentions Linden.
Through its embedded advanced analytics software engine, Ekkono has indeed become a force to reckon with—in the IoT space—and aims to become the de facto standard in machine learning for IoT. “We have also started coaching our clients, and it has turned out to be a common starting point to assist them and deliver a state-of-the-art solution,” concludes Linden.