Make the Connection Between AI and the IoT

May 7, 2019 Avnet

The Internet of Things (IoT) has been around for a very long time. In fact, the technology far predates the term. Some historians will tell you that the IoT moniker has been with us since the late 1990s, but we know that connecting our “things” started well before that. If you break IoT down to what it really means, it’s simply connecting the appropriate devices to the Internet to accomplish the required task. The ubiquity of the Internet and the proliferation of transceiver technology has greatly simplified that task.

The mass adoption of artificial intelligence (AI) is not nearly as long in the tooth, but it’s no spring chicken either. What’s new and very exciting is what happens when you put these two powerful technologies together. AI is as complex as IoT is simple. So, the marriage of the two can be a tricky endeavor. That “trickiness” is outlined in a recent article, Why Does It Take 10 Partners to Start an IoT Project? The bottom line is that the potential when combining AI and IoT comes from the insights that create efficiencies and/or increase profitability. That’s a big reason why having a partner helps ease the complexity so that you the engineer can focus on truly creating interesting and innovative technology.

A second term you should be aware of is machine learning. Machine learning is sometimes used interchangeably with AI, and that would be inaccurate. It’s actually a subset of AI, and much more focused, particularly in the industrial sector. An easy way to differentiate between the two is that where AI looks to achieve wisdom or intelligence, machine learning looks to acquire knowledge

Figure 1. The differences between AI, ML, and deep learning. Source: Exabeam.

AI lets you quickly pull insights from your data through analytics. Then, machine learning provides the ability to automatically identify patterns and detect anomalies in that data that edge-based devices like sensors and audio-visual systems generate. That information could be light, temperature, motion, pressure, humidity, air quality, vibration, sound, or video. Going one step further, speech or image recognition can help extract even more intelligence from the acquired data. Previously, pulling such information either required human intervention or simply was not possible.

Among other attributes, AI for the IoT lets manufacturers or other equipment users avoid unplanned downtime, increase operating efficiency, enable new and more complex end products, and enhance risk management. This comes from the insights offered by AI and machine learning, which allows not just for smarter machinery, but machinery that continually gets smarter over time.

Such downtime reduction comes thanks to predictive maintenance, which is using analytics to predict equipment failure ahead of time so you can schedule orderly maintenance procedures. Machine learning plays a big role here as it identifies patterns in the constant data streams coming from the equipment/machinery to predict equipment failure. Some estimates say that predictive maintenance can reduce the time required to plan maintenance by 20% to 50%, increase equipment uptime and availability by 10% to 20%, and reduce overall maintenance costs by 5% to 10%. An excellent whitepaper, Predictive Maintenance Use Cases in the Manufacturing Industry, outlines various predictive maintenance programs.

Thanks to vendors like Avnet, a lot of the hard work is done for you. Through its acquisition of Softweb Solutions, the company can handle most of the IoT software for you. You just need to connect it to your specific application. In addition, Avnet can also provide you with the data services to handle the analytics, which is why you wanted to connect to the IoT in the first place. Having that hardware-software integration tightens product development, reduces complexity, and gets the project completed in short order.

Avnet is of the belief that today’s AI-related software development requires out-of-the-box thinking, which is achieved in a number of ways. It starts with brainstorming and must encompass as many scenarios as possible that the platform may be exposed to. The more prior knowledge you can provide, the more accuracy you’ll see. Now you have help with the entire IoT solution life cycle, removing the need to research all the possible facets and variables because Avnet is a leader in both in both AI to IoT thanks to its long history that combines these disciplines.

Previous Article
A Funny Thing Happened on the Way …

As integrated circuit design progressed from the early days of manual layout to complex SoCs and fully auto...

Next Article
Experience the Power of AI with Solutions from AAEON

AAEON's range of AI solutions include compact embedded systems powered by NVIDIA Jetson TX2, and the innova...