Move Your AI to the Edge of the IoT
March 16, 2020
The advances made over the last few years in the field of AI are allowing this technology to permeate into all areas of industry, creating ?smart? applications and even smart industries.
The advances made over the last few years in the field of artificial intelligence (AI) are allowing this technology to permeate into all areas of industry, creating “smart” applications and even smart industries. This revolution has created the AIoT, which is a morphing of AI and IoT.
You could make an argument that the previous winners of the IoT explosion were the sensor makers and the Cloud providers. The latter were especially successful in that it was the keeper of the data—more data than most individuals could wrap their arms around. The next step, which was taken by most Cloud vendors, was to provide analytics to its customers: “You have all this data. Now you need to make use of it.”
But AI and the AIoT change the rules in that game. Now, decisions can be made at the Edge rather than in the Cloud. In most cases, the decisions are made quicker, because there is no delay because data remains at the Edge. And in many cases, the decisions are just as accurate as the Cloud-based decisions.
While AIoT has the potential to change the game, so to speak, it does not come without its challenges. These include the complexity associated with integrating new hardware and software, and the lack of engineering expertise in the space.
Aetina Corp. finds itself in the AIoT camp with the belief that the industrial sector is shifting its focus from the Cloud to the Edge. To that end, Aetina offers a rich and diverse product line of Edge computing devices that meet the needs of AI developers in the industrial space. And the company is clearly not alone. In fact, IDC estimates that annual Edge AI processor shipments will reach 1.5 billion units by 2023.
For example, Aetina offers GPU products like PCIe GPUs that are designed to accelerate graphics and empower modules that fit the embedded MXM form factor. In either case, the solutions could function as a system accelerator, a workstation that’s situated at the Edge, or a similar role.
1. The Aetina AN110 carrier board is designed around Nvidia’s Jetson platform.
Specifically, Aetina’s GPU-based computing platform series is designed around the Nvidia Jetson computing platform, with a series of carrier boards, including the AN110 (shown in Figure 1). The board’s small form-factor design allows them to be deployed on existing platforms where Edge-based AI can take performance to a completely different level.
The Aetina product line offers three key advantages in Edge-based systems:
- Privacy and security
- The ability to upgrade existing industrial equipment
- A proven business model
Hosting your data on-site is one way to ensure privacy and security. While the Cloud providers maintain that any data they receive can be kept secure, it’s that gray area that occurs between the Edge and the Cloud that can be vulnerable. The more nodes that exist, the more possible vulnerability points.
From a privacy perspective, intellectual property (IP) rights are as important as ever. Again, keeping the data on-site can’t guarantee privacy, but it does eliminate a series of potential vulnerabilities.
From an upgrade perspective, having a modular solution greatly reduces the required time and complexity. Industry experts claim that such an approach can actually reduce the time to add AI functionality by about 50%. This goes for both the hardware and software, as the upgrade is being performed on a proven platform with software already available.
Finally, Aetina has a long history of successful deployments in industrial graphics card and GPGPU solutions. The company has been providing its customers standard and custom products for more than a decade with full support and longevity.
To help developers get started down the correct path, Aetina has engineered a SparkBot project aimed at the AIoT. It’s currently based on Aetina’s Jetson platform series, and will likely migrate to other platforms in the future. The goal of the project is to reduce the overall system development time with features like the latest robot operating system (ROS), which is quickly gaining in popularity. The SparkBot project also allows for application concept sharing. In this scenario, Aetina would collect the application and operational ideas, then assist in building a prototype.
2. The TX2 Line Robot is designed with Aetina’s AN310 module and has integrated AI functionality.
An end product that’s based on one of the Aetina carriers, the AN310, is the TX2 Line Robot (shown in Figure 2). The robot is used on a manufacturing line to monitor progress. With its built-in AI functionality, it can pull in data, make some calculations, and make decisions which are then acted upon.
In summary, it’s clear that AI and IoT are merging in the industrial sector. Take advantage of the technology that is now at your disposal.