SWIM today announced general availability of its SWIM EDX edge software platform that provides analytics and machine learning to edge deployments.

SWIM CTO Simon Crosby explained that the platform ingests real-time edge data to construct an edge fabric. That edge fabric then applies the company’s analysis and machine learning capabilities to produce a “digital twin.” This twin is used to learn from the data coming into the system to find hidden patterns and predict future behavior.

“The problem we have targeted is processing vast amounts of edge data,” Crosby said. “Much of that data is murky, and you don’t know much about it. Our process is extremely efficient at the edge and can automatically learn from this murky edge data.”

(Crosby is also part of the company’s latest news as he recently joined SWIM. He previously served at CTO of Citrix’s data center and cloud division, and was co-founder of Bromium and XenSource.)

Digital twins are software models of physical items. This allows for the ability to model and monitor a physical item in real-time without actually touching the current status of that item.

According to a recent Gartner survey, and noted by ZDnet, 48 percent of companies in the Interet of Things (IoT) space are using, or plan on using, digital twins by the end of this year. The research firm added that more than 50 percent of manufacturers with at least $5 billion in annual revenues will have at least one digital twin initiative in place by 2020.

Crosby said many companies have approached analyzing edge data incorrectly by attempting to first move it to the cloud. He thinks this is flawed because analysis should be done close to where the data is collected for real-time insights.

“Everyone has suddenly figured out that the edge is fundamental,” Crosby half-joked. “They figured it out because it’s blatantly obvious. There is so much data at the edge that it’s preposterous to think you can upload all of it to the cloud to then be analyzed. For smart local decision making and to optimize an automated behavior you need to be able to make decisions at the edge.”

SWIM's managed software platform is designed with a small footprint, taking up two megabytes of storage space. This allows for integration into existing edge and IoT infrastructure. “Our goal is to get as close as we can to the edge. We don’t care how wimpy the computing power is, and we are compatible on very small systems,” Crosby said.

This also helps glean insight from edge deployments. “It’s really a problem of getting the fabric into the edge computing infrastructure out there,” Crosby said. “You need to be able to embed the agents on devices or add it as part of a management system.”

To solve this issue, SWIM is working on partner programs with IoT and edge equipment providers like Honeywell and Rockwell. "We are 100 percent focused on our partner programs," Crosby said. "This is a key component for us in gaining adoption from service providers, in the manufacturing space, and across government entities."

As an example of possible cost savings, Crosby cited the use of the SWIM platform to monitor traffic congestion in urban environments. He noted that to pipe data from each sensor up to a public cloud provider for analysis would cost around $50 per intersection per month. With SWIM’s edge-based analysis, Crosby said costs can be just one-tenth of that or less.

The cost analysis also plays into how a small company like SWIM expects to compete against more established players in the edge and IoT space.

For example, Amazon Web Services (AWS) last year launched its Greengrass platform that helps edge devices process data and communicate with the AWS cloud. The platform allows customers to use AWS Lambda to run code locally on connected devices similar to how they do it using AWS Cloud. Developers can also add Lambda functions to connected devices, and the devices can execute code locally and in real-time.

Crosby explained that companies like AWS are not appealing to developers because they require those developers to learn a different programming language.

“With Greengrass, Amazon is trying to attract developers to develop new applications to process the data from the system,” Crosby said. “My opinion is that there are not enough developers that understand the specific program language from either Amazon or the device maker to make that happen.”

Crosby said some of the SWIM software is open source, but that a majority is “bespoke and proprietary.”

“We have a program underway to foster developers and the environment, or basically a way to recruit from the developer community,” Crosby said, adding that it was “too early at this point to talk about that.”