CityScape

CityScape solves the problems of modern city planning including predicting the environmental impact of new developments, formulating efficient and proactive disaster response protocols, and others by utilizing big data, modern machine-learning and AI techniques, urban network analyses, and geospatial visualization tools. 

CityScape is a predictive geospatial visualization and simulation tool that utilizes the ever-growing corpus of data describing urban locales to show not only the current state of a city, but the probable impacts that urban planning decisions would have if carried out, drastically improving the urban planner’s ability to effect good decisions and the public’s ability to monitor those decisions.

If scaled globally, CityScape could improve the well-being of urban residents in terms of health, economy, and disaster readiness by influencing decision-makers with quantitative, data-driven analyses.

Urban populations are growing rapidly, with predictions of the percentage of the world living in cities exceeding 68% by 2050. This massive influx necessitates large-scale development and construction; since 2004 in New York City alone, 47 new skyscrapers have been built or are in development. Historical surges of urbanization often result in adverse impacts on urban environments as the pressure grows on urban planners to balance the interests of private developers with those of the people and their physical, emotional, and safety needs.

Urban planners have to contend with a multitude of factors that go into city planning initiatives and disaster response protocols, ranging from financial cost to zoning. Equally important and often harder to discern are the impacts that decisions might have on the health and well-being of the citizens. Poorly placed housing developments and highway systems, alterations to water supply and sanitation systems, and other common oversights in urban development can have devastating repercussions.

These decisions, if made correctly and with the proper influences, can greatly augment the well-being of urban residents and the ability of city governments and institutions to sustainably expand and meet the needs of the future.

CityScape is working to meet the needs of two sets of beneficiaries: city governments and other decision-making institutions, and urban citizens. The former group is concerned with the development and emergency response protocol decisions that take place on a daily basis within any city around the world, while the latter group is concerned with their individual and collective well-being and how it is affected by the former group’s decisions.

CityScape will address the needs of city governments and decision-makers by acting as a data-driven, objective advisor that can influence more well thought-out urban development and disaster response protocols. It will also address the needs of urban citizens by elevating their well-being by catalyzing sustainable, healthy urban planning decisions and by making urban planning decisions more transparent, as the tool will be available for their use as well.

Every city has its own unique composition and unique challenges, requiring unique solutions. CityScape will collaborate with both groups of beneficiaries in each city to source relevant local datasets and augment features that are required by that city. As conditions, needs, and data change, collaboration with local actors and commercial partners will be ongoing and comprehensive.

The solution

CityScape is a web application which visualizes urban planning. We aim to provide a one-stop-shop for urban planners and citizens who want to better make and/or understand development decisions as well as emergency response protocols.

Among other features is the user’s ability to quantify the effects of ongoing urban developments, commercial, residential, or otherwise, on their surrounding environment. Metrics shown include the predicted and ongoing effects on walkability in the surrounding area, the predicted environmental impact of the development, and additional effects based on the resolution of available data.

Several other features are demonstrated in this concept prototype. 

How it works

CityScape functions on a city-specific basis using data that precisely, temporally, and comprehensively describes that city, among which are population density by sector, traffic patterns, water main routing, localized pollution measurements, and socioeconomic distributions. These data are fitted to AI models that are used to create a virtual representation of the city. 

These AI models are what makes the magic happen: expanding the currently available geospatial visualization tools from mere representations of the past and present into powerful predictive engines that can forecast effects of decisions with objectivity heretofore unprecedented.

CityScape is based fundamentally upon existing open and closed source technologies, aiming to combine them in an innovative manner. Geospatial visualization (i.e. placing interactive objects on maps) is accomplished through tools like ArcGIS’s API. The fitting and application of AI models is accomplished through packages like Python’s SciKit-Learn. MIT’s own Urban Network Analysis Toolbox is utilized to quantify the physical layout of a city with accessibility metrics.

How it helps

There are two dimensions to the benefits of CityScape: those for urban planners, and those for urban citizens.

Urban planners will be able to make better-informed decisions that are able to manifest their good intentions for the well-being of the citizens they serve, without the pressure of having to somehow divine the possible effects of their decisions without aid.

Urban citizens will be able to hold their governments and other relevant institutions accountable by having access to the same decision-making data as those institutions and governments, enabling them to know when a development is being carried out for less-than-ideal reasons.

CityScape is innovative because of its novel combination of existing, well-established technologies to create a subtle, yet game-changing operational difference in geospatial visualization as it is known today: predictive mapping. The combination of geocoded data mapping, spatial analysis, AI modeling, and big data create a technology through which environmental consequences of urban developments can be both quantitatively predicted and visualized. CityScape bridges the gap between the back-end big data analyses and the user-friendly, front-end visualization of the results of those analyses.

This innovation is the logical next step in the GIS (geographic information systems) field, as big data is central to the ability to map descriptive data with both spatial and temporal precision. Technologies to predict the effects of decisions in a geographic context are just beginning to emerge as AI continues to spread through field upon field and industry upon industry, and CityScape is on the cutting edge of this wave.

CityScape's utilizes a couple of core technologies:

Geospatial Information Systems

The fundamental technology upon which CityScape relies is that of geospatial information systems. Included in this are services such as Esri's ArcGIS web API, which is a portal through which data produced and analyzed by CityScape can be visualized as an interactive web application for end-users.

Big Data

Without data, and lots of it, CityScape wouldn't be able to function reliably and with resolution and precision. Big data is the blanket term for a ballooning industry of mass data-collection, infrastructure-building, and distribution which allows modern applications that rely on large amounts of data, many of which are AI-related, to function. CityScape utilizes big data partnerships and infrastructure to consolidate relevant data describing urban locations, allowing for the training of well-fitted AI models to virtually simulate the city.

Artificial Intelligence/Machine Learning

The crown jewel of CityScape's technology, AI is what makes the magic of reliable prediction happen. CityScape uses machine learning (a practical implementation of AI) to create AI models that create a living, breathing simulation of a given city using a continuous flow of data describing that city. These AI models can be queried to predict the effects of changes in a city's fabric, offering data-driven, reliable, nuanced insights that would take a staff of many urban planners a prohibitive amount of time to deduce unaided.

Currently, CityScape is in prototype development, and is not yet serving a population. In one year, we plan to be in service of the Central Ohio area (where we are located), starting with lower-population semi-urban locations with populations of around 30,000-40,000 people. In five years, our goal is to serve urban centers throughout the American Midwest, with combined populations in the millions and even tens of millions.

Our goal within the next year is to continue to build the infrastructure and predictive-modelling back end of the CityScape system and apply this system to a range of pilot urban centers within our immediate geographic sphere of influence. This pilot will allow us to refine our service and work out any system-internal issues such as data bottlenecks while observing the practical impact of the service, which will serve to expose shortcomings and incubate further innovation in the system.

Within the next five years, our goal is to further develop and deploy this pilot into a reliable system that can be applied to any urban center with minimal end-user effort. In essence, our goal is to make CityScape as versatile and location-agnostic as possible within the constraints of AI modelling and available data.

Barriers for the next year:

• Acquiring standardized, precise, and high-resolution data describing the Columbus, OH area and the accompanying rights to this data: historically, the most difficult hurdle for a data-driven enterprise is the acquisition of the data needed to create the product. Cultural and political barriers exist to protect against the usage of data that could compromise individual privacy, and for good reason. One of CityScape's largest challenges will be continuously acquiring data with healthy public and private sector partnerships in a manner in which privacy is not intruded upon.
• Acquiring funding to research and develop well-fit AI models: it's no secret that the wrong AI model applied to a situation can yield completely unreliable and incorrect predictions. CityScape will need significant investment in order to research and develop AI models that are well-suited to represent urban environments.

Barriers for the next five years:

• Legal nuances for data usage in different cities, different states, and different nations: every locale has different laws governing its data policy and unique nuances in its partnerships with private sector services such as CityScape. These nuances will need to be worked through in order to satisfy the necessity for large amounts of high quality data, as CityScape won't be able to function in a satisfactory manner with any less.

3

The CityScape team is ideally skilled and situated to deliver our solution in order to make cities healthier. Each of our team members is educated in computer science, data analytics, and artificial intelligence at the university level and have experience in geospatial visualization and web-app development.

Being based in Columbus, Ohio is also a boon to our development capabilities because of the environment of innovation and forward-thinking in our city in relation to "smart city" projects such as CityScape. There are many local actors and institutions who would be interested in partnerships and collaborations in many aspects of the solution, from data sourcing to AI model creation to business model/process development.