I-Corps: RoadsInDB: Customer Discovery in the Logistics, Delivery, Ride Sharing, Location-based Services and Analytics Verticals

点击翻译按钮获取中文摘要

This project addresses logistics issues related to managing the deployment of large fleets of vehicles ranging from: Cab sharing services like Uber, Lyft, and Ola; food delivery services like Hub Grub; and retail delivery services like Google Shopping and Amazon, involving managing and optimizing the operation of hundreds and thousands of vehicles. Delivery networks were once dominated by a few large players and many small players. However, new models are evolving. For example, with 14k taxis on the streets of New York City (NYC), Uber is now is the single largest taxi service in NYC. Such companies must perform near real-time and rely on historical analytics on vehicles on road networks to keep their operations running efficiently. In particular, companies constantly compute network distances on road networks as they perform real time matchmaking between supply and demand. The technology involved in this project enables computing hundreds of thousands of network distances on road network at a very high throughput rate. This project's technology is different from Google Maps, Here Maps, and other online map services in the sense that they provide APIs to compute the shortest path between a source and destination vertex. This project's technology only focuses on computing network distances at a very high throughput inside a database, which makes this technology essential for companies that perform extensive analytical queries on road networks. This approach will be attractive to food delivery companies, package delivery companies (UPS, FedEx, USPS), taxi/limousine services, and ride sharing services (Uber, Lyft, Ola Cabs). Companies that provide map services such as Google, Microsoft, or ESRI may also find this technology attractive since using it they can improve their local search service by providing network distances instead of resorting to "as the crow flies" distances which is currently the case. The team's current business model would be to customize the system for each customer, requiring the purchase of licenses on a per-machine or per-site basis. This team's plan is to concentrate on several verticals (i.e., niche markets with specific, compelling needs) and understand their requirements. The team will develop applications to cater to each of the verticals.