Machine learning New deep learning algorithms are developing rapidly and can use data-driven design principles to solve historical problems. This is particularly exciting in the Internet of Things, where the rapid growth of connected devices has led to an explosive growth in the amount of data generated at the edge. These new algorithms play a key role in advancing the next phase of the IoT revolution.
There are many reasons why this method is adopted in many markets and fields, such as smart cities, smart homes, industrial Internet of Things (IIoT), wearable devices, etc. According to a market research company’s forecast, Internet of Things The market is expected to increase from US$5.1 billion in 2019 to US$16.2 billion in 2024, with a compound annual growth rate of 26%1).
A recent article published by PricewaterhouseCoopers (PwC) showcases the growth drivers of the Internet of Things and the benefits of artificial intelligence. Cost reduction is one of the benefits of AIoT. However, the proliferation of equipment and the increase in venture capital (VC) expenditures and the integration of information technology (IT) and operational technology (OT) with big data and cloud/fog are also occurring.
Since cost reduction is a benefit in several ways, it is not surprising that many system developers are interested in taking advantage of these combined functions in their next designs.In order to accelerate the development of differentiated AIoT products, Infineon Technologies released ModusToolbox™ Machine Learning (ML)The design tool supports deep learning-based workloads on Infineon’s PSoC™ microcontroller (MCU). ModusToolbox™ ML is a new feature ModusToolbox™ software and tools It provides designers with middleware, software libraries and special tools to evaluate and deploy ML models based on deep learning.
AIoT solves the system barriers in IoT. The original design concept of simply moving all data generated at the edge to the cloud for analysis and machine learning has encountered three basic obstacles: privacy, reliability, and latency. In order to reduce these obstacles, system designers changed the location of ML algorithms that are usually run on edge clouds. A good example is a voice-based smart assistant.
First, when you interact with an assistant, the time it takes to get answers back and forth is usually a bad user experience because it’s not a natural way for humans to interact. Secondly, the reliability and bandwidth of the Internet connection are also very important, especially when these assistants are running on wearable devices such as smart watches, you are not always able to connect to the cloud perfectly and reliably. Third, with these assistants everywhere, privacy is always the top priority, and trusting service providers with sensitive voice data is always a challenge. Running these algorithms efficiently at the edge removes these barriers and enables AIoT products to scale faster.
ModusToolbox™ ML allows developers to use their favorite deep learning frameworks, such as TensorFlow, and deploy them directly to PSoC™ MCUs. In addition, the tool helps designers by optimizing the model of the embedded platform by using various techniques (such as quantization) to reduce size and complexity.
ModusToolbox™ greatly simplifies the development of IoT products that use Wi-Fi and Bluetooth/Bluetooth low energy IoT products combined with RTOS system microcontrollers (such as PSoC™ series microcontrollers). Developers can use integrated middleware and code examples to easily connect their IoT products with leading cloud software platforms or proprietary cloud services.
In order to shorten AIoT development time, ModusToolbox™ also includes solutions that support popular ecosystems and cloud management tools, such as Pelion Cloud Management and Mbed™ OS, Amazon Web Services (AWS) IoT and FreeRTOS™ SDK, and Infineon AnyCloud IoT. In addition, it also provides specific tools such as low-power assistants, multi-radio smart coexistence, safety certification, and over-the-air updates to reduce the time and expense required to bring high-value, high-quality products to the market. This is the creation of future smart devices Provides flexible and easy-to-use tools and solutions.
One of the other key features brought by this tool set is to help system designers visualize how these optimization techniques affect model performance, so that they can make the right difference between performance and the size/complexity of the model running efficiently on PSoC™ MCUs trade off.
To help system designers get started quickly, code examples and IoT-focused development kits are provided to provide a smooth developer experience, thereby reducing the complexity that system developers face when developing AIoT applications. These usually require seamless machine learning workload integration, as well as computing, connectivity, and cloud domains. ModusToolbox™ ML can address these aspects by enabling developers to simplify their designs and put this functionality into any existing cloud or connectivity examples.
Figure 1. Optimization iteration is a key part of ModusToolbox ML.
ModusToolbox™ ML provides an unparalleled developer experience and reduces the complexity faced by system developers when developing AIoT applications. These applications usually require seamless machine learning workload integration, as well as the computing, connectivity, and cloud domains that ModusToolbox™ ML can solve.
ModusToolbox™ is available for download hereIn addition, code examples are shown in the ModusToolbox™ GitHub repository. To access online documentation, online videos, and regular real-time developer training, please join the Cypress Developer Community.
Click on here Learn more about Infineon’s machine learning solutions.
Click on Learn more about Infineon’s IoT solutions.
Do you like this story? Subscribe to the IoT business news!
The Internet of Things Business News provides electronic newsletters every half a month, each serving as an information resource to help business and technical professionals make effective design and business decisions.
Sign up to send news and updates to your inbox and read it anytime, anywhere.
