AI energy-efficient Edge processing targets the next generation of HMI

members Can Download this article in PDF format.

We have been immersed in traditional HMI methods for so long that adding artificial intelligence (AI) and edge computing requires fine-tuning our knowledge base. In the past, an HMI consisted of a physical control panel with pushbuttons, switches, trackballs, and indicator lights that enabled users to communicate with a machine. But it is not sacrilege to say that these people quickly leave the scene.

In its place, edge computing brings computing power to where your data is collected. Edge AI is smarter and more agile than other forms of data processing, as data is sent to remote data centers or the cloud for processing. It provides faster, localized processing with lower latency than traditional forms of cloud computing.

With the spread of artificial intelligence and machine learning (ML), the next generation of HMI is poised to offer entirely new ways of interacting with machines and devices. The coupling of cutting-edge AI features such as machine vision, analytics, and predictive maintenance with HMI applications is changing the way we interact with machines. Examples include gestures and/or facial recognition in a noisy factory environment or allowing devices to be controlled via digital voice assistants (such as Alexa and Siri) and a walkie-talkie.

Supported Resources:

Bringing HMI to the next level

As HMI continues to evolve, the processor technology behind these applications must be ready to enable this evolution. Recently, Texas Instruments released new high-end artificial intelligence processors targeting HMI applications. Designed for low power consumption, the first devices in TI’s Sitara AM62 processor family, the AM625 and AM623, are designed to provide power-efficient AI processing for dual-display and small form factor applications. It includes an AI accelerator that achieves high performance in deep learning for AI. Inference involves the process of analyzing data collected from an AI sensor and then making decisions based on it in real time or near real time.

TI has put great effort into developing the AM62x family (Chart 1). The processors consist of a 64-bit quad-core Arm Cortex-A53 microprocessor, a single-core Cortex-R5F microcontroller (MCU), and a Cortex-M4F MCU. While the AM623 targets IoT and gateway SoCs with Arm Cortex-A53-based object and gesture recognition, the AM625 specifically targets human-machine interactions with Arm Cortex-A53-based AI edge, and Full HD dual-display capability.

The best part may be the lower power consumption of these devices. AM62x processors achieve less than 500 MW of power consumption for a single A53 core operating at 1 GHz. About half of this has been achieved by similar low-power, low-cost processors in previous generations. AM62x processor power can be less than 1W when the quad A53 cores are running a cumbersome application on all four cores at 1.4GHz.

To bring the analytics to low-power high-end devices, including suspensions as low as 7MW without designing for thermal considerations, TI engineers used some clever techniques.

Low power design

Low power consumption is made possible by a simplified power architecture – the device features two dedicated power rails and five power modes. This device’s modular architecture delivers performance while supporting a variety of low-power modes without sacrificing critical system resources such as connectivity, power, security, safety, and cost.

Deep sleep mode at less than 5mW allows longer battery life, while active power under 1.5W is made possible with a base voltage of 0.75V. As a result, AM62 processors can reduce power consumption in industrial applications by up to 50% compared to competing devices, allowing an application powered by AA batteries to remain in operation for more than 1,000 hours.

Achieving optimal power performance is further simplified with the new TPS65219, a power management companion processor (PMIC) designed specifically to meet the power requirements of the AM62x processor. The device has an ambient temperature range of -40 to +105°C, making it suitable for various industrial applications. It includes three DC synchronous converters and four linear regulators.

Additional Features

AM62x processors include on-chip resources such as a universal asynchronous receiver/transmitter, multichannel SPI and I2C, and various connection options for common industrial sensors or controllers. These processors also provide dual Ethernet support and core EtherCAT support with a third-party ecosystem.

The parts also support a variety of display interfaces, including the RGB888 color format, where 8 bits are used for each of the red, green, and blue color components, and a low-voltage differential signal interface that supports 2K and full high-definition displays. The processors support dual displays, which provide engineers with design flexibility as two displays effectively increase the visual area and help make workflows smoother.

In addition, AM62x processors support multiple operating systems, including major Linux and Android operating systems.

The array of peripherals included in the AM62x enables system-wide connectivity, such as USB, MMC/SD cards, camera interface, CAN-FD, and GPMC (General Purpose Memory Controller) for a parallel host interface to an external ASIC/FPGA. The AM62x also supports secure boot for IP protection with an embedded Hardware Security Module (HSM) and advanced power management support for mobile and power-sensitive applications.

The AM625 and AM623 processors come in an ALW 13×13mm, 425-tooth package, with prices starting at under $5 for 1,000-unit quantities. The processors can meet the AEC-Q100 automotive standard in the 17.2 x 17.2 mm (AMC) package. Industrial and automotive functional safety requirements are met when integrated Cortex-M4F cores and dedicated peripherals are used, all of which can be isolated from the rest of the AM62x processor.


A rich hardware ecosystem, including a third-party evaluation module (EVM), helps designers work faster on application designs. SK-AM62 starter kit For HMI, IoT, and edge-based, AI is a low-cost standalone testing and development platform ($149) (Picture 2). It offers energy-efficient processing (featuring Cortex-A53, -R5F, and -M4F cores), full-resolution and dual-display support (HDMI and LVDO panels), software that includes edge AI demos, HMI, and more.


The AM62x’s low power modes and low active power consumption allow for a wide range of battery-powered uses, and the small form factor makes it easy to design the product without heat sinks or fans. Furthermore, 0.75V base voltage driver and power management features allow performance and power tuning for each application, resulting in a simple, low-cost power solution.

Supported Resources:

Leave a Comment