Embedded Systems and Artificial Intelligence are Connecting the Internet of Things



Artificial Intelligence – specifically computer vision and deep learning inference – are driving the evolution of IoT. 

Jonathan Luse, General Manager of IoT Planning and Product Line Management at Intel® Corporation

From sensor data analysis and aggregation to real-time customer engagement and targeted marketing, every company needs to have edge computing on its technology roadmap. When we talk about the edge, we are referring to a decentralized computing infrastructure in which data, compute, storage, and applications are distributed in the most logical, efficient place between the data source and the cloud. In general, edge computing moves data processing and storage resources closer to the source of the data. While many business decision-makers still consider “the edge” as exclusive to the network of devices embedded with electronics, software, and sensors that enable them to exchange and analyze data, commonly referred to as the Internet of Things (IoT); its value is becoming a cornerstone of enterprise data management at scale.

In fact, edge computing is as critical to delivering IoT-driven customer experiences today as content delivery networks were in the early days of the Web. These days, compute at the edge is facilitating real-time data processing that takes advantage of machine learning, deep learning, and computer vision in order to harness the data from billions of connected devices.

Powerful, energy-efficient computing at the edge, combined with the emergence of 5G bandwidth, is going to unlock the floodgates when it comes to connectivity. The number of IoT connected devices — from the user to the edge to the cloud — is astronomical. By 2025, 55.6 percent of all data will come from IoT devices, such as security cameras, industrial robots, digital signage, medical equipment, and other connected things.[1] AI is opening a world of new opportunities to extract value from this data, from monitoring quality control in smart factories to diagnosing health problems faster in hospitals. At least 40 percent of IoT-created data will be stored, analyzed, and acted on at the edge by 2019.[2]

The Value of Embedded Architecture

As the IoT evolves, small, high-powered computing devices with embedded processing capabilities will be at the confluence of performance and connectivity. Central to that functionality will be microarchitectures that facilitate better performance on mobile devices running at the edge. These system on a chip (SoC) platforms integrate both processor and field-programmable gate arrays (FPGAs) into a single device that can handle the rigorous demands of edge computing.

Moreover, for industrial applications, the functional safety (FuSA) and time deterministic capabilities embedded in Intel® processors are helping to ensure safe and timely execution of applications and workflows that accelerate the integration of robotics, security and control, and automation systems. These solutions provide end-users with both functional safety and high performance.

AI, and more specifically, computer vision and deep learning inference, are driving the evolution of IoT. The Intel® Distribution of OpenVINO™ toolkit, for example, helps companies, developers and research organizations to integrate these capabilities into their applications, and allows the flexibility to run them seamlessly across many types of Intel processors and accelerators, to deliver powerful data insights and results.

By 2025, 55.6 percent of all data will come from IoT devices.

The Intel® Distribution of OpenVINO™ toolkit, which includes a model optimizer and an inference engine, enables developers to streamline creating and deploying high-performance computer vision solutions using industry standard APIs, optimized frameworks, and libraries. The Intel® Distribution of OpenVINO toolkit calls target hardware plugins to accelerate deep learning workloads, taking advantage of all the “horsepower” inside a chip and providing dramatically improved AI inferencing performance boosts.

 

Accelerating vision computing with the Intel® Distribution of OpenVINO toolkit goes well beyond industrial applications. The smart cities, retail, healthcare, and transportation sectors will all benefit from the ability to process computer vision data at the edge to facilitate real-time decision-making. (See examples of real use cases.)

Edge compute solutions for emerging AI applications need to quickly and efficiently process large amounts of data. End-users need scalable solutions that can handle inferencing, general compute, storage, data security, and data, and analytics at the edge. Intel® Xeon® processors, for example, can also be configured with up to 20 cores to match the performance needs of edge systems, with the extensibility of up to eight processors on a platform.

Baked-in Security

As embedded devices become increasingly connected, developers should expect more than what has been historically characterized as stable, well-verified software equating with robust security. In today’s IoT environment, security has to start at the hardware level. That means security awareness begins even before a system is deployed.

When security is native on the hardware, data protection requires no processing overhead and has no impact on performance capabilities.

 

Of course, security remains a priority once a system goes online. Network protection has to focus heavily on security from top to bottom. That means securing data from its collection point through every step along the way as it is processed, analyzed, and stored.

When security is native on the hardware, data protection requires no processing overhead and has no impact on performance capabilities. In addition to a more secure system, overall costs will decrease, and performance will increase.

A recent report predicts that 45 percent of generated data will be processed, stored, and acted on at the edge by the end of next year, so security is a really big deal.

Designed for the Edge

Designing integrated systems for data processing at the edge is important for enterprise computing across industry sectors. It enables faster decision-making to happen when and where immediate action is needed. Data collection, management, and analytics at the edge also save bandwidth and computing time. For enterprise-level users, that means greater efficiency and lower costs.

Edge processing is also fast becoming de rigueur when it comes to delivering real-time customer engagement. Brand marketers need to be able to act on massive amounts of customer data efficiently and cost-effectively. AI and edge compute facilitate personalized engagement, and that translates into better customer experiences and increased conversions.

The fundamental value proposition when it comes to embedded technology is that it helps facilitate faster and more efficient data processing that can be used to inform real-time decision-making. Connected enterprises need integrated systems that can unleash the power of AI to keep track of a river of data flowing from the billions of connected devices during every second of every day.

Intel is helping power the IoT ecosystem with purpose-built silicon, interoperable systems, open architecture, and development support for everyone from OEMs and systems integrators to independent software vendors, developers and end users. Harnessing the power of connected data requires the processing power and bandwidth to ensure system integrity, from edge to the cloud—and everywhere in between.

At the center of the IoT universe are sophisticated, high-performing hardware and software systems that are tailor-made for connectivity and network security. Those are the embedded solutions that will connect the next generation of IoT devices.

Learn more about how embedded systems from Intel are connecting the Internet of Things at Booth 1- #338, at Embedded World 2019, in Nürnberg, Germany.

  1. IDC white paper, sponsored by Seagate, Data Age 2025, April 2017.
  2. IDC FutureScape: Worldwide Internet of Things 2017 Predictions.

 

 

Share and Enjoy:
  • Digg
  • Sphinn
  • del.icio.us
  • Facebook
  • Mixx
  • Google
  • TwitThis
Extension Media websites place cookies on your device to give you the best user experience. By using our websites, you agree to placement of these cookies and to our Privacy Policy. Please click here to accept.