Transportation

The Integrated Media System Center (IMSC) has a long history of applying novel approaches to solve real-world transportation problems. Under contract with the Los Angeles County Metropolitan Transportation Authority (LA-Metro), IMSC has been continuously collecting massive transportation datasets from LA County road networks since 2011 (more than 20,000 traffic sensors and GPS data from thousands of buses as well as thousands of incident reports collected by the California Highway Patrol). These real-world, raw datasets enabled fundamental research in traffic forecasting, game theoretic persuasion algorithms and control theory, which was funded by NSF, CalTrans, and several industry partners.

We have also invented new methods for high-rate streaming data retrieval, processing, storage, querying, analysis, and visualization using Microsoft StreamInsight, Oracle databases, and hierarchical multi-dimensional Data Cubes. Additionally, we have demonstrated the utility of “big” datasets (6 TB consisting of over 65 data elements) for planning research through several test applications. Because of our work, researchers from a variety of disciplines now have access to a rich repository of data ripe for analysis of complex transportation systems. For example, we teamed with the USC Sol Price School of Public Policy to answer questions such as, “Did the Expo Line increase transit patronage?” and “Did the Expo Line impact traffic performance?”

In a unique collaboration with the USC Annenberg School of Journalism, we enable data-driven journalism. Here we use our collection of LA Metro data to analyze traffic in Los Angeles and uncover underlying signals related to the economy, quality of life, safety, and social and economic status. Our findings include:

We have created the largest repository of transportation datasets in Southern California. Our end-to-end transportation data platform is being maintained and will be deployed at LA-Metro under two new contracts. In 2015, the proof of concept prototype of ADMS (Archived Data Management System) was successfully completed. Following the completion of the prototype, IMSC provided operations, maintenance, and technical services under a maintenance contract with LA-Metro.

In 2013, we started ClearPath, a company dedicated to building better navigation systems based on the concept of predictive path planning. ClearPath utilized several components of our patented IP in traffic forecasting and route planning. However, due to difficulties in penetrating the B2C market in navigation apps, we rebranded the company as TallyGo in 2015 with a new B2B business model.  TallyGo is the world’s first Navigation as a Service (NaaS) and empowers any company to embed a real-time navigation experience into their existing apps with virtually no development effort. If your business depends on moving people or things around, you can finally control and monetize the attention of customers and generate greater profit from your drivers. Significantly, you no longer need to depend on navigation controlled by the world’s largest companies. We believe navigation is for everyone. NaaS is the next revolution in SaaS — and TallyGo is the platform to accelerate that future.

Afsin Akdogan, Cyrus Shahabi, and Ugur Demiryurek, Throwaway spatial index structure for dynamic point data , Patent No. 9,501,509, United States, November 22, 2016

Bei Pan, Ugur Demiryurek, and Cyrus Shahabi, Traffic prediction using real-world transportation data, Patent No. 9,286,793, United States, March 15, 2016

Ugur Demiryurek and Cyrus Shahabi, Hierarchical and exact fastest path computation in time-dependent spatial networks, Patent No. 8,660,789, United States, February 2014

Ugur Demiryurek, Cyrus Shahabi, and Farnoush Banaei-Kashani, Efficient K-nearest neighbor search in time-dependent spatial networks, Patent No. 8,566,030, United States, October 2013

Giuliano, G., & Chakrabarti, S., Analyzing impacts of urban light rail investments: Study of the LA Metro Expo Line using archived real-time transportation system data; Transport Policy; Forthcoming.

Demiryurek, U., & Shahabi, C. (2017). Predictive Path Planning, Encyclopedia of GIS 2017, 1630-1640, 2017

Giuliano, G., Chakrabarti, S., & Rhoads, M., Using regional archived multi-modal transportation system data for policy analysis: A case study of the LA Metro Expo Line; Journal of Planning and Education Research, Published online September 24, 2015, doi: 10.1177/0739456X15604444 ; 2015.

Akdogan, A., Shahabi, C., & Demiryurek, U. (2016). D-ToSS: A Distributed Throwaway Spatial Index Structure for Dynamic Location Data. IEEE Transactions on Knowledge and Data Engineering, 28(9), 2334-2348.