Real-time data analytics and anomaly detection on a low powered (edge) device.
Problem: The amount of data is growing exponentially, mainly due to the rise of IoT, autonomous vehicles, drones, etc. Sending all this data to the cloud is not always acceptable or even possible. It adds extra latency, creates data bottlenecks (in low-bandwidth environments) or is simply not possible in distant fields where the connectivity is poor or expensive. We are coming to the point where we need to process huge amounts of data on the edge, low-powered devices with limited computing power, in order to gain accurate situation awareness in real-time Solution: SliceUp is a time-series analytics software that enables real-time decision making and failure detection on the edge (low powered) device in environments where there are very strict bandwidth or latency requirements. SliceUp continuously aggregates, fuses and analyses multiple data streams to alert human operators of any anomalies and send only important information to the cloud without flooding the network.
Zalety / korzyści z zastosowania technologii
Main benefits: 1) reduced latency for decision-making (for example in autonomous vehicles cannot accept the latency caused by round trips of data from the device to the cloud and back. They have to make decisions on the edge device). 2) optimized network bandwidth - with the amount of data generated every second by the edge devices, sending everything to the cloud can cause data bottlenecks. SliceUp enables data aggregations and analytics on the edge to send only important information to the cloud without flooding the network 3) reliability - in the distant fields where connectivity is poor or expensive you don't have to rely on it. Competitive advantage: 1) extreme speed - we are on average 100x faster than other time-series databases 2) co-deployment (edge+cloud) - other time series databases cannot be deployed on the edge device - it would require re-writing the whole code 3) AI-assisted decision making - It can detect anomalies in the absence of vast amounts of historical training data 4) UX focused - simple deployment under 1 minute and easy and intuitive UI that doesn't require data science skills (graphical user interface - GUI).