Developing Intelligence for Smart Crowd-Sourced Urban Delivery (CROUD) System The underlying thesis of the CRowd-sourced Urban Delivery (CROUD) technology is the ability, enabled by recent advances in communication and ubiquitous mobile computing , to match highly fragmented transport capacities with vastly diverse demand for urban deliveries, temporally, spatially and in real-time. While CROUD is currently being deployed in various forms, many aspects of this technology have not yet been rigorously studied. To claim a niche market in which it can thrive and even dominate in the long run, the existing CROUD platforms need rigorous business analytics and operational tools and to more effectively engage humans – namely shippers (i.e. consumers) and carriers (i.e. couriers) – in the system. The overarching goal of this project is to enhance and expand the capabilities of the CROUD technology so as to provide a smarter, faster, and greener solution to the last-mile delivery puzzle that is challenging not only transportation professionals but indeed the entire retail industry. To this end, a new Partnership For Innovations (PFI) is proposed to forge collaboration between a CROUD-based technology firm, namely Zipments, and two research universities, Northwestern University (NU) and University of Illinois at Chicago (UIC). The proposed partnership will bring together Zipments’ technical group (including its CEO and co-founders) and five researchers with diverse expertise to develop intelligence that aims to transform existing CROUD platforms into smart service systems. Specifically, the envisioned smart CROUD system will be built on four interconnected subsystems that encapsulate, respectively, pricing/matching mechanisms, consumer/courier management strategies, collaborative delivery/routing algorithms and real-time data collection and analysis tools. Prototypes of these new systems will be implemented and evaluated on Zipments’ CROUD platform. Intellectual merits This project is the first to systematically investigate the problem of designing, managing and operating a smarter CROUD system. First, it will create pricing strategies that can adapt to varying market conditions through innovative matching mechanisms. Analyzing pricing with matching models is a novel approach in a two-sided market such as encountered in CROUD systems, and it promises insights beyond markets considered herein. Second, the project will collect new behavioral data and develop econometric models that can be used to guide the ways by which a CROUD system interacts with its consumers and couriers. These models can also help understand and predict human choices in a CROUD system, a field completely new to behavioral econometrists. Third, the proposed research will develop smartphone-based motion activity detection methods that are uniquely suited to track couriers. Among other things, these methods can recognize higher level activities such as parking and reroute, minimize energy consumption of the device, and fuse indicators from multiple sensors. A database design and a prototype domain-specific query language specific to the data collected in this project will also be developed. Last but not least, the project will develop computational tools for real-time collaborative delivery/routing. In doing so, it will contribute new formulations, algorithms and insights to challenging problems in operations research and computational economics such as combinatorial auction, network design with relays, real-time vehicle routing considering transfer, het
|Effective start/end date||9/1/15 → 8/31/20|
- National Science Foundation (IIP-1534138)
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.