Telit Cinterion and Alif Semiconductor have introduced Vision AppKit, a revolutionary connected camera reference design that sets a new standard for intelligence and energy efficiency in the market.
This compact camera system demonstrates the integration of advanced Edge AI capabilities with secure and efficient wireless communication technologies.
The microcontroller (MCU) in the Vision AppKit is the Ensemble E3 from Alif Semiconductor. It combines 2 Arm Cortex-M55 real-time MCU cores and 2 Arm Ethos-U55 micro-neural processing units (microNPUs). Alif has developed chips with security features. The aiPM technology reduces power consumption and extends battery life by powering only the active areas of the chip. For example, a smart IoT device can turn off unused MCU or microNPU cores. This combination speeds up AI device tasks such as face/object recognition and image classification, while consuming 2 orders of magnitude less power than other edge AI processors.
Two Telit Cinterion modules provide wireless communication. The WE310F54 module supports Wi-Fi 4 (b/g/n) and Bluetooth Low Energy 5.0. And the ME310 module provides cellular access to LTE CAT-M networks. Both modules have power-saving features. Since the modules are pin-compatible for the Vision AppKit reference design, either module can be selected.
The Vision AppKit utilizes the Onsemi MT9M114 digital image sensor with 720p resolution. Using the dimensions of the WE310 module as a reference, we estimate that the board measures approximately 31 x 38 millimeters. It also features a USB-C interface and the ability to connect a LiPo battery.
The microcontroller (MCU) in the Vision AppKit is the Ensemble E3 from Alif Semiconductor. It combines 2 Arm Cortex-M55 real-time MCU cores and 2 Arm Ethos-U55 micro-neural processing units (microNPUs). Alif has developed chips with security features. The aiPM technology reduces power consumption and extends battery life by powering only the active areas of the chip. For example, a smart IoT device can turn off unused MCU or microNPU cores. This combination speeds up AI device tasks such as face/object recognition and image classification, while consuming 2 orders of magnitude less power than other edge AI processors.
Two Telit Cinterion modules provide wireless communication. The WE310F54 module supports Wi-Fi 4 (b/g/n) and Bluetooth Low Energy 5.0. And the ME310 module provides cellular access to LTE CAT-M networks. Both modules have power-saving features. Since the modules are pin-compatible for the Vision AppKit reference design, either module can be selected.
The Vision AppKit utilizes the Onsemi MT9M114 digital image sensor with 720p resolution. Using the dimensions of the WE310 module as a reference, we estimate that the board measures approximately 31 x 38 millimeters. It also features a USB-C interface and the ability to connect a LiPo battery.