IBM this week announced it was open-sourcing Quarks, a very interesting technology that enables organizations to analyze Internet of things (IoT) data locally, on gateways or at edge devices.
According to IBM, it is far more costly and time consuming to extract meaning from IoT data by transmitting over a network and then analyzing it via a centralized application, than to do the analysis at the edge.
However, working in conjunction with centralized analytic systems, Quarks enables a common streaming analytic model across different kinds of devices using familiar tools, IBM said on a web page describing the technology. It runs locally analyzing the continuous streams of data coming from equipment, vehicles, systems, appliances, and sensors of all kinds, IBM said.
Quarks is an open source programming model and runtime for edge devices that enables users to analyze data and events at the device. By doing analysis at the edge, you can reduce the amount of data that you transmit to your analytics server as well as reduce the amount of data that you store.
A Quarks application uses analytics to determine when data needs to be sent to a back-end system for further analysis, action, or storage. For example, you can use Quarks to determine whether a system is running outside of normal parameters, such as an engine that is running too hot, the Quarks website said.
Moreover, Quarks enables you to shift from sending a continuous flow of trivial data to the server to sending only essential and meaningful data as it occurs. This is especially important when the cost of communication is high, such as when using a cellular network to transmit data, or when bandwidth is limited, the website said.
Quarks can be embedded in gateways and devices and can be used in conjunction with vendor and open-source data and analytics solutions such as Apache Kafka, Spark and Storm. Quarks is API driven and modular.
“This builds on a long-running strength that IBM has had since it bought Rational many years ago, namely a focus on the ecosystem of device developers and ISVs,” Al Hilwa, an analyst with IDC, told eWEEK. “We are seeing many major players make a play for the IoT developer, which is proving a hot area as we march faster and faster towards the digital world. We have seen interesting solutions from many players like Cisco, Salesforce and Microsoft around processing IoT data. What IBM can do really well is connect the world of devices with the world of data and cognitive processing in unique ways and bring that technology to its digitally transforming clients.”
IBM Open-Sources Quarks IoT Service to Its Own Gain
Quarks is a micro-kernel style runtime for execution enabling analytics on small-footprint edge devices or sensors, such as Raspberry Pi devices or smart phones. It has Java APIs for developing applications that execute analytics using a per-event streaming paradigm. It also features connectors for MQTT, HTTP, JDBC, Apache Kafka and the IBM Watson IoT Platform. And it features multi-platform support including Java 8, Java 7 and Android. Producing a stream that contains a phone’s sensor events is an example of Android-specific functionality, IBM said.
“Given the diverse nature of edge devices, local analytics is best addressed by an open community of software experts and device experts committed to expanding the capabilities and real-world use of Quarks,” IBM said.
IBM is encouraging the community to get involved in expanding the capabilities of Quarks.
IBM’s decision to donate its Quarks technologies to open source is the latest in a long line of open source investments by the company, including Eclipse and its Power processor architecture,” said Charles King, principal analyst at Pund-IT. IBM hopes the platform will gain momentum via the contributions of energetic open source developers. But the company also stands to gain if Quarks achieves a level of market momentum that eventually enables IBM to build commercial offerings around it.
Indeed, King said he believes there is a very good chance of Quarks doing just that.
“As IoT has evolved, a strong consensus has emerged around the need to develop solutions that support ‘intelligence’ — in the form of compute/analysis capabilities — at the edge of the network in endpoint and gateway devices, he said. “Doing so would allow those solutions to initially determine relevant data which could then be transmitted back to central data centers for closer analysis, thus saving time and communications costs. Quarks is a toolset designed to do just that and if it catches on, IBM could be in the catbird’s seat in terms of Quarks expertise and service development.”
Long time IBM watcher and founder of the Enderle Group, Rob Enderle sees possibly an even bigger role for Quarks.
“Quarks is their streaming platform for processing lots of data at once and IoT has ‘lots of data at once’ as one of its big requirements,” Enderle said. “In addition, the one huge sustaining advantage of their System z architecture is massive I/O giving them a potential significant competitive advantage with IoT if it uses an architecture that System z can best use. Quarks is that platform and, if successful, it could create a significant advantage for System z hardware for large scale IoT deployments.”