Sponsored

Edge AI Tech Can Lead Healthcare Industry Through the Quarantine

Written by Aetina

Innovation and rapid technology development continue to drive improvements in medical treatment and healthcare. And the rising star among such technologies is artificial intelligence (AI). AI has proven its worth in all industries, and medical is no exception. AI platforms employed as edge computing solutions provide the advantages of lower latency, easy deployment and security—smoothing the design process for system developers.

AI as Part of National Healthcare Policies Worldwide
After proving its value in markets such as factory automation, AI has now expanded to all industries. Recognizing the value of AI, governments worldwide have been making significant investments in the application of AI to healthcare in their respective countries. On the other hand, from the side of medical fields, what/how the worldwide government has done proofed the influences and importance of AI to healthcare as well.

Here are some examples: The Japanese government highlights AI in healthcare in its Society 5.0 effort. The UK Health Secretary announced £250 million investment AI. The US government continues to apply AI into their healthcare system, and, last April, the United States FDA authorized the IDx-DR , an AI-driven software could detect signs of diabetic retinopathy in retinal images. The Taiwan MOST (Ministry of Science and Technology) reported on its achievements in project AI innovation of medical imaging.

Clearly, AI is becoming entrenched as a leading technology in the medical field. AI-based machine learning is used to train data for cancer analysis, AI-assisted medical imaging and big data computing for public health analysis and strategies. Meanwhile, AI-controlled healthcare robots help supplement the workforces of today’s hospitals.

For its part, Aetina is well positioned in its role as an AI power transformer. Medical is a professional industry comprised of many specialized fields. In contrast, AI is a general technology that’s applicable to almost any situation. The power of AI can feed the needs of any vertical application. Aetina offers a rich and diverse product line of edge computing devices designed to meet the needs of AI developers in the industrial space. Meanwhile, earlier this year at Embedded World 2020, Aetina launched a plan called SparkBot, including three experimental robots that feature the smart logistic concept. Aetina is a NVIDIA Jetson preferred partner, so the first three prototypes are based on Jetson computing modules. Those are combined with Aetina-created carrier boards—each designed to serve the AIoT market. The core value of the SparkBot plan revolves around the idea of generalizing edge AI computing applications.

Aetina demonstrated one of their robots called Surface at Embedded World 2020. It’s built using the concepts of autonomous vehicles such as detecting its environment and avoiding danger.

From Cloud to Edge, Emerging AI Applications Show Their Value
There’s a valid argument to be made that the winners in AI traditionally play in the cloud computing side of the technology. But as issues like business security and communication latency have moved front and center, the emphasis has shifted to toward edge computing. For its part, market analyst firm IDC estimates that annual Edge AI processor shipments will reach 1.5 billion units by 2023. Aetina is focusing on the edge computing, and by doing pre-integration it provides a robust environment for AI ideas, with a focus on doing machine learning work on specific platforms appropriate for an application. Beyond that, Aetina also develops experimental robots, not only for demonstrations, but also as a means to implement intelligent applications.

While AI is critical to many markets and industries, it has the potential to change the rules at a business level. Moreover, during the current COVID-19 pandemic quarantine period, AI can empower the capacity of medical institutions and be a big help to the decision-making processes of healthcare authorities worldwide.

Aetina has developed three experimental robots based on three different NVIDIA Jetson-based Platforms, including AN110-NAO, AN310-TX2and AX710-XAV. Made by NVIDIA, the computing modules are GPU-based with high performance and low power consumption. Meanwhile, the carrier boards, made by Aetina, ignite the power of AI. This combination of technologies offers a solution that’s extremely relevant to medical applications, particularly during the pandemic quarantine.

Along just those lines, Aetina has expanded the concept of Smart Medical technology with its SparkBot line of experimental robots. The Dot robot is powered by Aetina’s AN110-NAO platform, based on the NVIDIA Jetson Nano module. The Line robot embeds the AN310-TX2 platform, based on the Jetson TX2 module. And the Surface robot runs the company’s AX710-XAV platform, based on the Jetson AGX Xavier module.

Figure 1: Aetina SparkBot Dot

AI Benefits Healthcare Industry
Now, let’s look at each robot in more detail. The Dot (Figure 1) is the smallest robot of the SparkBot series. It’s well suited to serve as a healthcare gate entrance monitor. To serve those needs, Aetina integrated an air sensor and face-mask-detection software into the platform. Once it’s set up at the entrance of a hospital, Dot could record the air condition and stop people from entering without a face mask.

Going a step further with this concept is Aetina’s customer SmartCow. SmartCow offers a platform called Edgar, an AI-powered healthcare assistant.  SmartCow’s product does face-mask recognition with the anti-coronavirus multi-functions algorithm running on Aetina’s AN110-Nano platform. Edgar comes with six different AI models: social distance estimation, masks, gloves detection, medical apron detection, medical full-body suit detection and safe sneeze action classification—all very beneficial functions to offload from healthcare personnel.

Figure 2: Aetina SparkBot Line

Next, the Line robot (Figure 2) serves as a nursing robot. The unit is equipped clip, depth camera and skeleton detection software. In the case of some infectious diseases, patients are unapproachable because of the dangers to healthcare workers. This puts nurses tasked with the caring responsibilities at great risk. A care robot like Line can offload many unsafe tasks from nurses. The Line robot could hand stuff to the patients and move by easy gesture control. Meanwhile, skeleton detection functionality can be used to make sure that patients stay safe on their beds.

Figure 3: Aetina SparkBot Surface

Finally, the Surface robot (Figure 3) provides the highest AI performance in the series. It is an advanced service robot for the healthcare environment. Surface is designed using concepts of autonomous vehicles. That enables Surface to move around by itself, using environmental detection and perception to prevent any accident or collision. In a hospital setting, Surface could be used to transport medication and biomedical waste through the hospital, lowering the chance of unnecessary touch from human to human.

Industry Connection is Key
As a general technology, applicable across all fields and markets, AI is enriching the possibilities in countless vertical markets. But no industry impacts human beings as directly the medical/healthcare space. Where once all decision making and medical tasks relied solely on medical practitioners, now AI technology can assist and offload many critical medical functions. Moving forward, such AI will not only save lives but also provide much needed support to healthcare organizations worldwide.

About Aetina Corporation
Aetina was founded in Taiwan in 2012, as a long-term support provider of high-performance GPGPU and Jetson edge AI computing solutions for embedded applications. We are focusing on the industrial market, providing industrial components and longevity service. In addition, we’re also an integrator in the AIoT market, helping customers by offering smart, innovative and reliable GPU solutions that serve the needs of the AIoT market and solve their complex problems.

Content provided by Aetina


Don't miss out on upcoming issues of Circuit Cellar. Subscribe today!

 
 
Note: We’ve made the October 2017 issue of Circuit Cellar available as a free sample issue. In it, you’ll find a rich variety of the kinds of articles and information that exemplify a typical issue of the current magazine.


Would you like to write for Circuit Cellar? We are always accepting articles/posts from the technical community. Get in touch with us and let's discuss your ideas.