Industrial Internet of things (IIoT) is often referred to as Industry 4.0: a major upgrade of today’s industrial operations. Real-time analytics, data value and operational decisions are key goals for Operations Technology (OT) deployed within the industry.
However, for industrial applications to scale, stay connected, be secure and process data quickly and efficiently, intelligent data processing needs to happen at the network edge.
It was clear to the founders of IoT database management system (DBMS) company Machbase back in 2013 that machine or sensor data is critical for enabling IIoT real-time analytics – however, this results in a data tsunami of massive tables and high-speed ingestion. In order for the application to be effective, a database engine that is incredibly fast and works at the industrial device endpoint, such as a robot in a smart factory, is needed.
Machbase has become that database and is on a mission to enable the full potential of data insight on the IIoT edge. It has produced the fastest time-series DBMS, with lofty aims to be the go-to database for all future IIoT operations.
The Arm Blueprint team recently sat down with Machbase CEO Andrew Sungjin Kim to find out more about how the company’s Arm-based technology is transforming the future of IIoT with advanced edge compute for industrial applications.
What does success look like at Machbase?
We enable IIoT companies to improve their operations, increase efficiencies, eliminate downtime, and achieve higher ROI. We currently enable operations in many applications including smart logistics, agriculture, smart home, on the grid, in smart cities and smart buildings, mining and manufacturing to make sense of their data in real-time at the edge and on the cloud.
Success for us is when we save our customers money and time. As an example of what this looks like at our company, we recently improved a factory logistics operation that needed to collect and process 100,000 records a second per node across 500 nodes. To put that into context, that’s 50,000,000 records a second on which to perform vibration testing and anomaly detection. This was performed at the edge and aggregated on an on-premise server. The result was a vast improvement in uptime of the factory floor, with no manufacturing shutdown, saving millions of dollars in downtime.
Robotic applications on a production line cannot afford unexpected wear, cracks, or particles to interrupt rail performance or cause damage – and this requires the real-time analytics at the edge that Machbase provides.
How is Machbase currently using Arm in its solutions?
Machbase offers a complete unified product platform that includes three products: Edge, Fog and Cluster/Cloud editions. It works on the entire spectrum of Arm products and their licensees, including many edge examples, such as nVidia Jetson, Raspberry Pi 3, and ARTik.
Arm’s low-power, high performance technology enables edge computing for us. Machbase is able to run 100s of databases on Arm processors, from processing on smart factory floors to a utility’s distribution infrastructure to an automotive ecosystem. We enable decentralized and distributed data analysis through a wide area of system architectures, leveraging Arm’s offering at the network edge and fog, and preparing data for big data analytics in the back-office cloud/cluster.
Arm provides Machbase with scalable compute solutions that provide flexibility to choose the best fit in terms of power and form factor, allowing Machbase to deploy in volume and in ways previously unimaginable.
What do you think is going to be the most significant change for the industry in the next five years?
I believe edge compute is the next disruptive phase in the computing evolution, similar to what desktop compute did to central computing in the 80s.
A few drivers of edge compute include real-time response, security and safety concerns against DDoS attacks, cost reductions and the need for artificial intelligence (AI) and machine learning (ML) close to the source.
What are you doing next to solve the industry’s most significant change?
Machbase now has a robust database platform for all of IIoT. Next, Machbase will offer better visualization of the data from years to microseconds in real-time.
Machbase sees better integration with the cloud technologies and the direct support for artificial intelligence and machine learning at the edge itself.