SBIR Phase I: Computing-Assisted Zoning Optimization and Service

这个小企业创新研究（SBIR）第一阶段项目是建立基础，开发一个计算辅助重划产品。重新划分或分区是将空间划分为地区或区域的过程，同时在各种空间限制下优化一套标准。重划是一项极具挑战性的任务，在许多现实问题中通常需要重划，例如政治/学校重划，紧急区域设计，服务管理和业务规划。商业市场上现有的重新划分工具都采用手动方法，依赖于用户以试错的方式优化各种标准。这些手动工具效率低下，使用起来非常耗时。该项目将基于新开发的重划优化算法和可视化控件构建一个原型重划产品，该产品比现有算法更快，优化质量更上级，并首次提供了以自动或半自动方式满足最终用户需求的可能性。拟议项目将（1）设计一个灵活的架构，以系统地将各种优化标准集成到新算法中，以满足不同的需求;（2）显着提高处理大型数据集的计算可扩展性;（3）通过实际应用评估原型产品。这个项目更广泛的影响/商业潜力是，它可以从根本上改变重划选区的方式在实践中进行。学术界现有的自动化重划方法不能满足实际需要，这是市场上没有采用自动化方法的产品的原因。现有的商业工具基于手动方法，完成重划任务通常需要一名或多名全职技术人员一次工作数月。通过从该项目创建产品，即使是非技术用户也可以在几分钟内生成专业计划。潜在客户包括政府（市、县）、学区、提供重新划分服务的咨询公司、需要优化服务交付的私营企业，甚至有兴趣参与重新划分决策的普通公民。该产品面向美国和国际市场，包括公共和私营部门。

This Small Business Innovation Research (SBIR) Phase I project is to establish the foundation for developing a computation-assisted redistricting product. Redistricting or zoning is the process of dividing space into districts or zones while optimizing a set of criteria under various spatial constraints. Redistricting is an extremely challenging task that is commonly needed in many real-world problems, such as political/school redistricting, emergency zone design, service management, and business planning. Existing redistricting tools in the commercial market all employ a manual approach, which relies on the user to optimize various criteria in a trial-and-error manner. These manual tools are ineffective and extremely time-consuming to use. This project will build a prototype redistricting product based on a newly developed redistricting optimization algorithm and visual controls, which is much faster and more superior in optimization quality than existing algorithms, and for the first time offers the potential to meet end user requirements in an automated or semi-automated manner. The proposed project will (1) design a flexible architecture to systematically integrate a variety of optimization criteria in the new algorithm to address different needs; (2) significantly improve the computational scalability for handling large data set; and (3) evaluate the prototype product with real-world applications. The broader impact/commercial potential of this project is that it could fundamentally change the way redistricting is performed in practice. Existing automated redistricting methods reported in academia do not meet practical needs, which is the reason that there is no product in the market that employs automated methods. With existing commercial tools, which are based on a manual approach, completing a redistricting task usually requires one or more full-time technicians working for months at a time. With the product being created from this project, even a non-technical user may generate a professional plan within minutes. Potential clients include governments (cities and counties), school districts, consulting firms who provide redistricting services, private businesses that need to optimize service delivery, and even general citizens who are interested in participating in redistricting decision making. The product targets the U.S. and international market, including both public and private sectors.