SIROM3 -- A Scalable Intelligent Roaming Multi-modal Multi-sensor Framework

Abstract

Understanding the future transportation infrastructure performance demands a smart Cyber-Physical Systems (CPS) approach integrating heterogeneous sensors, versatile computing systems, and mobile agents. However, due to sensor versatility and computing intricacy, designing such systems faces challenges of immense complexity in mobile sensor fusion, big data handling, system scalability, and integration. This paper introduces SIROM 3 , a Scalable Intelligent ROaming Multi-Modal Multi-Sensor framework, for next generation transportation infrastructure performance inspection. SIROM 3 offers a scalable and expandable framework through orthogonally abstracting software / hardware structures in a layered Run-Time Environment (RTE), which facilities sensor fusion, distributed computing, communication and mobile services. A Heterogeneous Stream File-system Overlay (HSFO) and a flexible plugin system (PLEX) are embedded in SIROM 3 to simplify big data storage, processing, and correlation. To evaluate the scalability of SIROM 3 , we implemented a mobile sensing system of 30 heterogeneous sensors and 5 computing platforms coordinated by 1 data center. SIROM 3 's expandability is highlighted by adding an advanced radar platform which required less than 50 lines of C++ code for integration. Over 20 terabytes of data covering 300 miles have been collected, aggregated, and fused using SIROM 3 for comprehending the pavement dynamics of the entire city of Brockton, MA. SIROM 3 offers a unified solution and ideal research platform for rapid, intelligent and comprehensive evaluation of tomorrow's transportation infrastructure performance using heterogeneous systems.Current roadway pavement monitoring methodologies often face challenges such as intrusive data gathering (e.g. stopping traffic), manual efforts and subsequently infrequent data collection and limited coverage [33]. Hence, non-intrusive, automated, fast, and adaptive solutions for data collection and infrastructure assessment are necessary. Heterogeneous sensor systems such as Multi-Modal Multi-Sensor (MMMS) systems are promising aiming for adaptability, automated operations, power and fuel efficiency, and ubiquitous assessment capability in multiple data domains [9], [12]. Integrating MMMS systems onto a mobile platform creates a Roaming Multi-Modal MultiSensor (RMMMS) system to collect multi-modal data under roaming conditions. However, RMMMS systems are challenging to develop and operate due to the heterogeneity in sensors, data types and synchronization principles, as well as the sheer number of sensors. Typically sensor systems are designed tailor-made to a specific application, which impedes the overall scalability. Meanwhile, system complexity increases exponentially with computational diversities, network-wide collaboration and data correlation. In addition, data-intensive sensors produce a large volume of streaming data in real-time raising the importance to effectively store, access, and process big data. Additionally, deploying multiple RMMMS to increase geogr...

Framework and implementation of a continuous network-wide health monitoring system for roadways

“…Significance of automation lies in the nature of the project: VOTERS system would need to collect hundred Gbytes of data a day. Here, we will briefly discuss VOTERS system integration design and achievements, more details can be found in [16][17][18]. …”

“…The integrated sensing system has been deployed on the VOTERS prototype vehicle and an on-board controller manages the positioning and the above listed sensor systems as shown in Figure 2. The on-board controller manages the individual sensor systems, synchronizes data streams with sub-ms accuracy, registers all data streams in time and space, and an access point to the control center [16][17][18].…”