 I travel to China frequently because Carnegie Mellon has a partnership with Sun Yat Sen University in Guangzhou in the areas of engineering and brain research. I'm very excited to be here in Tianjin to present our research and have a discussion about smart infrastructure. As you know, business leaders have a number of tools in place to monitor the market performance of their enterprise. When a business starts to get into trouble, as was seen in the 35 millimeter film industry, there are many established indicators that help these leaders identify the problems and make corrections. Not taking action to change the business model, however, will often lead to failure. In the US alone, trillions of dollars have been invested in civil infrastructure in bridges, in roads, in water systems, and sewers. The American Society of Civil Engineers finds that most of this infrastructure is in poor condition and estimates that would take over two trillion dollars to improve this situation. Unfortunately, for much of our existing civil infrastructure, we don't yet have the right data collection tools in place to be able to efficiently and effectively monitor that infrastructure. The data that we do have is manually collected every one or two years, oftentimes based on human visual inspection, and at times loaded with errors. I'm here to talk about my field of smart infrastructure, which is a blend of build infrastructure such as bridges, with many network sensors collecting data over time and space, and sophisticated data analytics applied to be able to predict and visualize the condition of that infrastructure. In other words, smart infrastructure provides those early indicators of problems that would allow a decision maker to act in a very timely and decisive way in the way that a business in trouble would also be dealt with. The goal is to better manage infrastructure, saving money and energy, and reducing or eliminating disruptions to businesses in the public. To do this, we need to collect sensor-based data, such as loads, strains, and vibrations, and apply sophisticated analytics to that data. Using the data and analytics, we're able to build models that will help us better understand the conditions to which that infrastructure is exposed, any concerning trends in its behavior, and options for remediation. However, placing and maintaining sensors in an external environment can be a very challenging endeavor. A Carnegie Mellon research project is exploring using vehicles as bridge sensors. The idea is that vehicles will carry power and protect the set of accelerometers, and then transfer that data for analysis. This will yield lower installation and maintenance costs for hundreds of thousands of bridges. We apply sophisticated data analytics to this vehicle to be able to recognize problems with the bridge, such as excessive stress, loss of material due to corrosion, or loss of structural support. We've tested this concept in a laboratory using a model vehicle and a model bridge, and are able to expose it to different damage scenarios and test different sensors. The analytic techniques that we're using are similar to the signal processing and image analysis used to detect cancers, tumors, and CAT scans. We've demonstrated it's possible to use this analytics to detect subtle changes made to the bridge model using only the data collected from the model vehicle. Outside the laboratory, we've been able to install accelerometers on in-service public transit vehicles. The data from these vehicles is used to monitor bridges and the concrete railroad ties over which the vehicles travel, detecting problems much sooner, and creating a much more efficient inspection process. Stepping back from the bridge context, what are some of the challenges of delivering smart infrastructure? Well, first, we need to secure smart infrastructure in the same way we need to secure many other cyber systems. We need infrastructure information models that help us better organize the collected data, and third, we need new business models that will help municipalities affordably adopt smart infrastructure. So let's think about what the future holds. Imagine a smart water main that knows that it's going to break, basically turns off the valves in the right order, schedules a maintenance crew, avoids collateral damage in the area, and saves hundreds of thousands of dollars that might otherwise be spent in an emergency repair. But smarter infrastructure systems should allow us to have more cost-effective management, more efficient operation, and much more reliable performance. Two types of businesses will likely emerge that will collect this data and provide third party services. And finally, insurance companies will hopefully offer greatly reduced premiums for these smart infrastructure systems. I hope this talk has provided some insight into the potential impacts of smart infrastructure. However, the existence of this technology alone will not allow governments to have an impact if they cannot affordably adopt it. So my questions for this discussion period are, what business models might allow municipalities to take up smart infrastructure in the future, and what will motivate them to do so? Thank you.