How Open-Source Software Drives IoT and AI
How Open-Source Software Drives IoT and AI
One of the most promising emerging developments is the intersection of the IoT and AI. Expect more of this as open source continues to speed development in these exciting technologies.
Join the DZone community and get the full member experience.Join For Free
Hortonworks Sandbox for HDP and HDF is your chance to get started on learning, developing, testing and trying out new features. Each download comes preconfigured with interactive tutorials, sample data and developments from the Apache community.
Twenty years ago, open source was merely a new buzz phrase. Some scoffed at it, many misunderstood it, and only a small subset of people believed it could change the world. Today, open-source software lies at the heart of the most exciting technology developments.
The term may go back to the late '90s, but open sharing of software source code dates back even further to the '50s, when mainframe companies shared operating system source code with others. The movement rapidly gathered force with the success of the World Wide Web and now drives a significant portion of the economy. Open-source software has touched everything from word processing applications to databases and has recently enjoyed success in a major growth area: the Internet of Things (IoT). The movement has proven useful in both the development of edge-based IoT equipment and the evolution of back-end analytics systems that make sense of the data.
Connection and Interoperability
At the network's edge, one of the biggest challenges facing manufacturers is interoperability. The IoT hit the mainstream before the evolution of standards to ensure connectivity between the various manufacturers' devices. In this massively fragmented landscape, open source software can create a cohesive mechanism for data exchange. The community can work together to create connectors that allow data to flow across multivendor IoT networks.
Initiatives such as the Linux Foundation's EdgeX Foundry and Apache NiFi/MiNiFi are good examples of these efforts. The open source protocol and secure flow management provide the foundation for different vendors' equipment to exchange information in industrial IoT environments, speeding up the deployment of interoperable IoT infrastructures for customers.
Speed of Innovation
In the back end, organizations must process and analyze the ocean of data flowing from IoT infrastructures — which, in many cases, contain tens of thousands of devices. Open source is a key factor in the development of this software.
By relying on open source code bases, organizations like Apache and their associated communities can accelerate the development of analytics tools on scale-out hardware architectures, enabling them to map and reduce the IoT's large data sets.
Not only does this allow the community to quickly refine analytics capabilities for IoT data types but it also makes it possible to process that data across different cloud infrastructures. It's easier to flip your on-premises workload up to a public cloud infrastructure provider if the same code base is available in both domains. Commonly accepted open source code bases make that more likely.
Open source is also helping companies to keep pace with artificial intelligence (AI) as it evolves at a breakneck pace. In 2015, Google gave its deep learning framework, TensorFlow, to the open source community, and it has since become a common tool for people creating deep learning training models for AI applications.
TensorFlow it is not the only AI framework to hit the open source community. Theano, developed at the University of Montreal, is an open source project for machine learning using Python. Torch, a framework that Facebook developed with others including Twitter, is an open source project that spawned a Python version called PyTorch. And Microsoft, once a harsh critic of the open source movement, has since warmed up to the concept. In 2016, it released the AI framework behind its Cortana digital assistant as open source software called the Microsoft Cognitive Toolkit.
Enterprise-Grade Open Source
Why are companies in such a competitive field releasing the source code for all to see? It's because a rising tide carries all boats. By making their code visible to everyone, they also make it possible for people to improve. This accelerates software innovation-not only in the base platform, but also in specialist versions of the software used in niche industry applications.
In addition, open source offers a platform for governance. Open intellectual property enables companies to put collaborative checks and balances in place that improve security in structured, reliable ways. One example of this is the ODPi governance initiative. The program, which is governed by the Linux Foundation, aims to standardize and simplify the big data ecosystem. Governance frameworks such as this program make open source tools stronger and more shareable.
A Future Driven by Open Source
One of the most promising emerging developments is the intersection of the IoT and AI. We're already seeing this as image recognition finds its way into webcams at the consumer level and industrial equipment makes its own preventative maintenance decisions. Expect more of this as open source continues to speed development in these exciting technologies. The ideal scenario is one in which open source, AI, and IoT create a cohesive whole that is more than the sum of its parts.
Published at DZone with permission of Piet Loubser , DZone MVB. See the original article here.
Opinions expressed by DZone contributors are their own.