Developing an AI IoT application with open software on a RISC-V SoC

Abstract

RISC-V is an emergent architecture that is gaining strength in low-power IoT applications. The stabilization of the architectural extensions and the start of commercialization of RISC-V based SOCs, like the Kendryte K210, raises the question of whether this open standard will facilitate the development of applications in specific markets or not.In this paper we evaluate the development environments, the toolchain, the debugging processes related to the Sipeed MAIX Go development board, as well as the standalone SDK and the Micropython port for the Kendryte K210. The training pipeline for the built-in convolutional neural network accelerator, with support for Tiny YOLO v2, has also been studied. In order to evaluate all the above aspects in depth, two low-cost, low-power, IoT edge applications based on AI have been developed. The first one is capable of recognizing movement in a house and autonomously identify whether it was caused by a human or by a house pet, like for example a dog or a cat. In the context of the current COVID-19 pandemic, the second application is capable of labeling whether a pedestrian is wearing a face mask or not, doing real-time object recognition at a mean rate of 13 FPS. Throughout the process, we can conclude that, despite the potential of the hardware and its excellent performance/cost ratio, the documentation for developers is scarce, the development environments are in low maturity levels, and the debugging processes are sometimes nonexistent.

A Survey of the RISC-V Architecture Software Support

“…The work of [31] uses a Universal Sensor Platform (USeP) SoC to meet the requirements imposed by IoT, which features a RISC-V processor and integrates a range of peripherals with a scalable subsystem as a Three Dimensional System-in-Package (3D-SiP). Further, the Sipeed MAIX-Go development board employed by [66] features a Kendryte SoC K210, with a dual-core 64-bit RISC-V with the RV64GC ISA, including an Floating-Point Unit (FPU) compliant with support to single and double-precision multiply, divide, and square-root operations.…”

“…Torres-Snchez et al [66] evaluate the development environment toolchains and debugging process concerning the Sipeed MAIX Go development board and Tiny YOLO v2 support by deploying a low-power IoT edge application, achieving good performance and cost characteristics.…”

Edge Analytics for Bearing Fault Diagnosis Based on Convolution Neural Network

“…There are a number of works using edge AI devices for various practical applications, such as face mask detection [9], resilient image compression for IoT cameras [10], mineral granulometric analysis [11], conveyor belt longitudinal rip detection [12]. Most of the work focuses on the problems of image analysis.…”

Intelligent Security Monitoring System Based on RISC-V SoC

“…Bailey et al [16] demonstrated a signal analysis SoC consisting of a general-purpose RISC-V core with vector extensions and a fixed-function signal-processing accelerator. Enrique et al [17] used open-source software to develop AI IoT applications on RISC-V SoC. The advantages of the RISC-V architecture in security and power consumption promote the application of RISC-V in the Internet of Things (IoT), mobile terminal, edge computing, and other fields.…”