There is a huge opportunity to reduce costs, improve efficiency, and improve safety through acting on insights from the wealth of new data at the edge. But how? Enterprises can’t afford to ship all this raw data to a remote cloud, find a data scientists and develop analytics to identify insights and actions. Nor can they afford to train new staff to build, deploy and manage complex new on-prem apps, sensors, networks, storage, and compute clusters. And they cannot risk a security breach.
What’s needed is an edge-based “plug & play” software product that securely and affordably delivers the powerful benefits of Insights, Intelligence to brownfield environments, without the complexity and cost of bandwidth, storage and cloud based processing, and without requiring new skill-sets or management frameworks. What’s needed is a product that quickly delivers insights directly to operators, staff and apps to facilitate fast decision making - without complex configuration, integration and testing.
SWIM was built to address this need: A powerful, affordable edge-based addition to data-rich environments that transforms rich data into intelligent, responsive, business-optimized resources using powerful edge machine learning. SWIM software is designed to “plug and play” in existing industrial environments, where it autonomously weaves a resilient, secure, distributed fabric that delivers context-relevant insights precisely where and when they are needed. SWIM consumes data on inexpensive devices at the network edge, intelligently transforming floods of low-value events into a manageable stream of learned insights and predictions that help operators and applications quickly respond.
SWIM learns from its environment. Its goal is to learn powerful insights for timely delivery to operators, business executives and IT applications - with minimal configuration or ongoing management.
SWIM is designed to drop into data-rich environments and find its feet fast. The product requires minimal configuration – before diving into your data to hunt for unseen trends and anomalies. SWIM nodes need to know how to find each other and where to report data. They need to be told what protocols to listen on, and how to make API calls to business systems such as ERP platforms. Configuration is simple and can be done by existing staff or SWIM. Then SWIM takes over and does the rest. Crucially, if there is ever a problem with device hardware, it is straightforward to just swap in a new device, and let SWIM figure out how to securely configure and provision it.
SWIM automatically hunts for insights in the flood of real-time data generated by typical complex environments - a torrent of continuous data packets from sensors in production facilities, traffic data on roads, and sensors in smart utility networks that must be aggregated, analyzed, and reported on. SWIM identifies correlated data in your environment without configuration, consuming vast amounts of low-level event data on commodity devices, at the edge, saving you from the need to transmit, store and process it in the cloud.
SWIM aims to protect the enterprise from attacks from four critical vectors, by building verifiable security into the system by design.
SWIM security starts with an immutable hardware root of trust to ensure that the system will always bootstrap into a known-good initial state. It also protects SWIM from manual attacks by physical access, and it enables each SWIM node to stamp a certificate of provenance on every piece of data so that applications can reason about end-to-end trusted data flows.
SWIM introspects its own behavior and learns about its typical behavior so that any anomalies can be immediately reported.
SWIM stands as sentinel between the hazardous (and potentially compromised) environment, and enterprise IT systems. It learns and predicts the behavior of sensors and if they behave abnormally, can immediately alert and silence malicious inputs.
Similarly, SWIM protects the environment from potential attacks from the IT domain and the internet.
SWIM embraces the acceleration of the price/performance curve in computing, delivering real-time, edge-based solution that uses commodity devices to learn incomparable insights on the fly, avoiding the cost and complexity of cloud based transport, storage and processing.
SWIM, which aims to revolutionize environments with a simple, resilient, secure and self-managing solution that automatically learns and predicts key performance indicators. SWIM is designed to add value fast, without extensive POCs, and without disturbing existing environments and tools. SWIM runs on commodity hardware, at the network edge, making projects cheap to pilot and affordable to scale solutions for:
SWIM: Powerful, affordable edge-based addition to data-rich environments that transforms complex environments into intelligent, responsive resources.