A few years ago, no one would have believed that a global team of volunteers could write a major operating system kernel or a comprehensive encyclopedia. But today, Linux and Wikipedia and many similar open source projects serve as evidence that loosely organized teams can create and maintain the complex technical and intellectual infrastructure upon which society increasingly depends. These projects have succeeded not by accepting participants indiscriminately, but by creating a meritocracy that rewards expertise regardless of its source. Yet interestingly, what is viewed as the best cognitive model (the crowd), has been applied to relatively non-technical societal problems, with its full capacity potentially untapped. This project adopts the same open source philosophy to establish new modalities of research, teaching and practice responsible for our nation's civil infrastructure.
This domain is naturally suited towards open source concepts: complex systems like buildings, bridges and other lifelines (e.g., power, water) have many different stakeholders, require many kinds of technical expertise, and have a lifetime far beyond any one contributor, yet impact the lives of many within society as a whole. As the information used in the design, analysis and even management of civil infrastructure is trapped inside proprietary systems, these essential projects often do not benefit from the full expertise and latest advances available in the wider engineering community. To address this problem, a virtual organization (VO) is being created: Open Sourcing the Design of Civil Infrastructure (OSD-CI) (Fig. 1). This VO allows stakeholders – engineers, public officials, researchers, students, and even the public at-large – to review, propose, modify, evaluate, and contribute to designs, analyses, databases and other fundamental research that supports civil infrastructure and provides these Citizen Engineers with unparalleled access to tools that will empower their individual initiatives. By leveraging the collective knowledge, pooling computational resources, archiving project data, and accelerating research to practice, the challenges facing our civil infrastructure can be addressed in an unprecedented fashion.
The Civil Engineering Challenge
The investment in civil infrastructure is second only to health care in this country, representing 10% of the gross domestic product and arguably serving as the backbone of the nation’s economy. While the current economic stimulus acknowledges the need to quickly address its deteriorating condition, the research and design practice dedicated to it still relies on limited resources scattered physically throughout universities, governments, private laboratories within various industries, design offices, and trade organizations with little formal collaboration. Thus, the analyses employed are often simplified, not only for convenience and practicality, but also due to a lack of access to recent technological advances in the field and complete knowledge of the complex interactions inherent to these systems. This leads to conservatism at the expense of efficiency or even worse, simplicity at the expense of structural integrity and safety. Clearly, today’s challenges to civil infrastructure are far too complex and multi-disciplinary to be addressed by yesterday’s simplified, compartmentalized philosophy, which has stymied collaborative thinking and innovative means of discovery.
The Open Source Computing Challenge
There are many successful examples of open sharing using computer systems. Systems such as Condor and SETI@Home have demonstrated how to share computer hardware in order to attack very large problems that exceed that capacity of any one machine. Projects like Linux and Apache have shown how to share source code in order to progressively construct large software systems that increase in quality and capability over time. Projects like Wikipedia have shown how to share knowledge in order to construct and refine a public encyclopedia. Systems like Amazon's Mechanical Turk do the opposite by using computer systems to harness human effort on tasks where computers fail, such as identifying images. While some projects have demonstrated how collaborating users may share computing hardware, and others have demonstrated how to share software, share knowledge, or share human effort, to our knowledge, no existing system enables collaboration in all four dimensions. OSD-CI does so, with the end goal of supporting the design of civil infrastructure. The computer science (CS) challenge lies in making each dimension of collaboration effectively reflect and exploit the social and technical constraints of the other dimensions. While this in and of itself is challenging, it is further compounded by the fact that these open source, social computing principles have never been applied in high risk application domains where failure may have serious implications. Thus, the OSD-CI Collaboratory presents a transformative challenge in computer science: invent a social computing framework that embodies these open source principles across four dimensions of sharing, while satisfying the social and technical needs of the civil engineering (CE) community, especially the need for both open and extremely trustworthy processes to ensure participation without compromising quality.