Transportation research is carried out across the various IBM labs. The role of the Smarter Transportation PIC is to stimulate research and collaboration within the IBM Research community for the study of theory, methods and applications in transportation management, optimization and control.
Areas of interest include but are not limited to,
- Transportation systems for passengers and freight
- Multimodal transportation: air, maritime, rail, urban transportation
- Intelligent mobility information services for citizens
- Transportation network simulation & modeling
- Traffic prediction and flow control
- Human mobility, travel behavior and network economics
- Sustainable and innovative transportation solutions
- Multimodal dynamic routing/scheduling
At IBM Research - Australia transportation is seen as a component of complex real-life problems that we encounter and we address it in the context of the main research directions of the lab. In the case of disaster management, transportation is a key component of evacuation, emergency response services, but also of the recovery process. Similarly, the exploitation of natural resources relies on efficient transportation, often multi-modal (trucks, trains, ships). In the context of IBM's Smarter Cities, we consider urban transportation: in recent years traffic congestion has become an issue that needs to be addressed.
The Smarter Transportation research team at the IBM Research - China focuses on developing analytics and simulation tools to better understand and optimize the urban transit and traffic network. We have developed TOPS(Transit Network & Operation Optimization System) improving urban transit services around three areas:(a)Passenger Demand Pattern Discovery: leveraging boarding/alighting counts to gain an insight in how, when and where passengers travel. Using this information, the (time-varying) passenger demand can be estimated and predicted, and then the transport resource capacity (e.g.,bus network;transit terminal layout) can be accordingly optimized for both offline planning and real-time operation;(b)Transit Network Assessment & Optimization: developing business analytics and optimization to ensure and maintain the transit network desired level of quality of service to best accommodate the passenger demand both at an aggregated level and disaggregated to each individual passenger;(c)Multimodal Transit Simulation: Developing the multi-modal simulation tool to assess various transit management plans, help turn predictive insights into operational reality, close the gap between strategy and execution, and help decision maker survive and thrive in these challenging conditions.We have developed TIPS(Traffic Insights Platform & Services) aiming to help cities to build a sustainable and reliable transport network. The TIPS solutions are around three areas:(a)Heterogeneous Data Fusion for Large Volume Traffic Data:applying data mining technologies to effective use of periodicity, local predictability, as well as statistical information offline or real-time; (b)Traffic Congestion Pattern Analysis and Prediction: to reveal traffic congestion tempo-spatial patterns and improve congestion prediction capability under normal and abnormal conditions (e.g., adverse weather, incident);(c)High Performance Map-Matching of Trajectory Data: able to process hundreds of millions of records on a regular PC server,with paralleling processing capability
Transportation research in Dublin is performed by the Smarter Urban Dynamics team that focuses on developing analytics and tools to better understand and optimize urban dynamics. The team develops research around three areas. (a) Leveraging digital traces from mobile devices to gain an insight in why, how and when we travel. Using this information, mobility demand can be predicted and transport resource capacity accordingly allocated, offline or in real-time. (b) Leveraging transport infrastructure and associated applications in ITS analytics to ensure and maintain the desired level of quality of service to best accommodate the mobility demand both at a global level and down to individual travellers such that demand and supply are better matched. (c) Providing travellers with an awareness of the transport infrastructure conditions, so that each can individually and confidently decide for a best travel strategy that is also collectively sustainable.
IBM Research - India is focused on technologies which lead to increased adoption of Intelligent Transportation Systems (ITS) in developing countries. We have questioned the traffic problem which is relevant to be solved, investigated sensing techniques that work in the aggregate (using people, audio and Call-Data Record) since fine-grained sensing is infeasible due to chaotic traffic, and proposed dynamic journey finding for multi-modal public transportation with no sensors since that is the reality on the ground as well as how car-pooling should be promoted by businesses. Our unique focus within IBM has been on data issues around transportation, since that influences what algorithms become applicable. We drive this all the way to formal modeling and standards. See here for details.
Beyond developing countries, the team has experience in advanced cities like Boston, USA where it participated in a Smarter Cities Challenge around data aggregation with state-of-art sensor data. The team did a well-received traffic tutorial at AAAI 2012 whose slides are here
The Singapore Collaboratory has a team dedicated to working on the next generation of transportation technologies in partnership with the Singapore Land Transport Authority. Ongoing projects include the Minimum Sufficient Network problem which addresses optimal transport sensor placement, and the Decongestion Engine, which in collaboration with the team in WRC aims to provide a comprehensive suite of transportation command center decision support tools for managing unexpected and planned events on the network.
IBM Research - Tokyo focuses on large-scale microspic traffic simulator called Megaffic (IBM Mega Traffic Simulator). Megaffic is an agent-based simulator of traffic flows with two unique features built on an X10-based agent simulation middleware called XAXIS. First, Megaffic can build its model of simulation by directly estimating some of the parameters of the model from probe-car data. This capability is in contrast to existing agent-based simulators of traffic flows, where the values of their parameters are calibrated with iterative simulation. Second, Megaffic can run on massively parallel computers and simulate the microscopic traffic flows in the scale of an arbitrary city in the world or even the whole Japan. Most of the latest supercomputers use massively parallel architectures, where large number of multi-core processors and GPGPUs are interconnected. The massively parallel architectures have brought new challenges to simulator developers in handling large-scale agent-based simulations. Developers, who are often unfamiliar with parallel computing, have to be aware of the underlying multi-node architectures to take full advantage of their potential. In addition, they must pay attention to the difficult task of resource allocation to the individual nodes, which is especially critical in large-scale agent-based simulations with millions of agents. To comply with the massively parallel architectures, we developed X10-based Agents eXecutive Infrastructure for Simulation (XAXIS), a new platform for ultra-large agent-based simulations on massively parallel supercomputers. XAXIS is built on top of X10, a new parallel programming language suitable for multi-core architectures, which is being developed by IBM Research.
The Watson Research Center (WRC) includes several teams focused on creating innovative Smarter Transportation solutions based on a variety of approaches and use cases. One area of considerable activity is Real-Time Transportation Analytics, which includes the work that has led to the Traffic Prediction component of the Intelligent Operations Center (IOC) (formerly called Traffic Prediction Tool, or TPT) as well as real-time traffic state estimation (called Data Expansion Algorithm). The team's current focus is the research, development and piloting of Decision Support System Optimizer (DSSO) capability in various cities around the world. The DSSO is a comprehensive suite of transportation command center decision support tools for managing unexpected and planned events on the network. WRC also includes the City in Motion team who leverage cell phone records to estimate origin-destination demands for travel over a transport network with an aim to the re-optimization of public transport (e.g. bus) routes.