Smart Wireless Data Transfer Solution for Underwater Sensors Powered by Edge Computing

Abstract

Abstract With the ever-increasing number of underwater assets and structures driven by the energy industry's shift to renewables, the need for reliable long-term remote monitoring solutions for both environments and assets is growing. Such solutions typically consist of a suite of sensors, the data transfer mechanism, and data analytics and visualization. Underwater wireless technology is the key enabler for such a holistic monitoring solution that can be used for a variety of applications. However, many commercial operations are designed around the assumption that underwater wireless technology is not yet suitable for today's operational needs. While there are several advancements in the domain of underwater wireless systems in recent times, they have not made it to operational commercial systems. In this paper, we look at the example of Acoustic Doppler Current Profilers (ADCPs), where a typical operation relies on either offline data download or using a cable to perform real-time data transfer. We then illustrate how such an operation can be transformed by integrating a smart modem that packs technologies such as software-defined design, edge computing, and machine learning along with robust and reliable high-speed underwater wireless communications to an ADCP to achieve flexible and reliable wireless data transfer. The software-defined design aspect allows easy integration to a variety of sensors, in this example an ADCP. The edge computing and machine learning aspects allow the modems to optimize the data for a high-speed acoustic link that supports adaptive modulation and automatic retransmission to avoid data loss. Smart scheduling techniques are incorporated into the solution to support extremely low-power modes to enable long-term deployments. An application-specific web-based user interface (UI) provides a seamless user experience during the whole operation. Integrating such innovations to form a holistic solution for long-term monitoring can drive down overall costs and improve operational safety.