KwangSoo Yang

Spatial big-data challenges intersecting mobility and cloud computing

Increasingly, location-aware datasets are of a size, variety, and update rate that exceeds the capability of spatial computing technologies. This paper addresses the emerging challenges posed by such datasets, which we call Spatial Big Data (SBD). SBD examples include trajectories of cellphones and GPS devices, vehicle engine measurements, temporally detailed road maps, etc. SBD has the potential to transform society via next-generation routing services such as eco-routing. However, the envisaged SBD-based next-generation routing services pose several significant challenges for current routing techniques. SBD magnifies the impact of partial information and ambiguity of traditional routing queries specified by a start location and an end location. In addition, SBD challenges the assumption that a single algorithm utilizing a specific dataset is appropriate for all situations. The tremendous diversity of SBD sources substantially increases the diversity of solution methods. Newer algorithms may emerge as new SBD becomes available, creating the need for a flexible architecture to rapidly integrate new datasets and associated algorithms.

Experiences with evacuation route planning algorithms

Efficient tools are needed to identify routes and schedules to evacuate affected populations to safety in the event of natural disasters. Hurricane Rita and the recent tsunami revealed limitations of traditional approaches to provide emergency preparedness for evacuees and to predict the effects of evacuation route planning (ERP). Challenges arise during evacuations due to the spread of people over space and time and the multiple paths that can be taken to reach them; key assumptions such as stationary ranking of alternative routes and optimal substructure are violated in such situations. Algorithms for ERP were first developed by researchers in operations research and transportation science. However, these proved to have high computational complexity and did not scale well to large problems. Over the last decade, we developed a different approach, namely the Capacity Constrained Route Planner (CCRP), which generalizes shortest path algorithms by honoring capacity constraints and the spread of people over space and time. The CCRP uses time-aggregated graphs to reduce storage overhead and increase computational efficiency. Experimental evaluation and field use in Twin Cities Homeland Security scenarios demonstrated that CCRP is faster, more scalable, and easier to use than previous techniques. We also propose a novel scalable algorithm that exploits the spatial structure of transportation networks to accelerate routing algorithms for large network datasets. We evaluated our new approach for large-scale networks around downtown Minneapolis and riverside areas. This article summarizes experiences and lessons learned during the last decade in ERP and relates these to Professor Goodchilds contributions.

A critical-time-point approach to all-start-time lagrangian shortest paths: a summary of results

Given a spatio-temporal network, a source, a destination, and a start-time interval, the All-start-time Lagrangian Shortest Paths (ALSP) problem determines a path set which includes the shortest path for every start time in the given interval. ALSP is important for critical societal applications related to air travel, road travel, and other spatiotemporal networks. However, ALSP is computationally challenging due to the non-stationary ranking of the candidate paths, meaning that a candidate path which is optimal for one start time may not be optimal for others. Determining a shortest path for each start-time leads to redundant computations across consecutive start times sharing a common solution. The proposed approach reduces this redundancy by determining the critical time points at which an optimal path may change. Theoretical analysis and experimental results show that this approach performs better than naive approaches particularly when there are few critical time points.

A Lagrangian approach for storage of spatio-temporal network datasets: a summary of results

Given a set of operators and a spatio-temporal network, the goal of the Storing Spatio-Temporal Networks (SSTN) problem is to produce an efficient data storage method that minimizes disk I/O access costs. Storing and accessing spatio-temporal networks is increasingly important in many societal applications such as transportation management and emergency planning. This problem is challenging due to strains on traditional adjacency list representations when storing temporal attribute values from the sizable increase in length of the time-series. Current approaches for the SSTN problem focus on orthogonal partitioning (e.g., snapshot, longitudinal, etc.), which may produce excessive I/O costs when performing traversal-based spatio-temporal network queries (e.g., route evaluation, arrival time prediction, etc) due to the desired nodes not being allocated to a common page. We propose a Lagrangian-Connectivity Partitioning (LCP) technique to efficiently store and access spatio-temporal networks that utilizes the interaction between nodes and edges in a network. Experimental evaluation using the Minneapolis, MN road network showed that LCP outperforms traditional orthogonal approaches.

Benchmarking Spatial Big Data

Increasingly, location-aware datasets are of a size, variety, and update rate that exceeds the capability of spatial computing technologies. This paper addresses the emerging challenges posed by such datasets, which we call Spatial Big Data SBD. SBD examples include trajectories of cell-phones and GPS devices, vehicle engine measurements, temporally detailed road maps, etc. SBD has the potential to transform society via a number of new technologies including next-generation routing services. However, the envisaged SBD-based services pose several significant challenges for current spatial computing techniques. SBD magnifies the impact of partial information and ambiguity of traditional routing queries specified by a start location and an end location. In addition, SBD challenges the assumption that a single algorithm utilizing a specific dataset is appropriate for all situations. The tremendous diversity of SBD sources substantially increases the diversity of solution methods. Newer algorithms may emerge as new SBD becomes available, creating the need for a flexible architecture to rapidly integrate new datasets and associated algorithms. To quantify the performance of these new algorithms, new benchmarks are needed that focus on these spatial big datasets to ensure proper comparisons across techniques.

A Critical-Time-Point Approach to All-Departure-Time Lagrangian Shortest Paths

Given a spatio-temporal network, a source, a destination, and a desired departure time interval, the All-departure-time Lagrangian Shortest Paths (ALSP) problem determines a set which includes the shortest path for every departure time in the given interval. ALSP is important for critical societal applications such as eco-routing. However, ALSP is computationally challenging due to the non-stationary ranking of the candidate paths across distinct departure-times. Current related work for reducing the redundant work, across consecutive departure-times sharing a common solution, exploits only partial information e.g., the earliest feasible arrival time of a path. In contrast, our approach uses all available information, e.g., the entire time series of arrival times for all departure-times. This allows elimination of all knowable redundant computation based on complete information available at hand. We operationalize this idea through the concept of critical-time-points (CTP), i.e., departure-times before which ranking among candidate paths cannot change. In our preliminary work, we proposed a CTP based forward search strategy. In this paper, we propose a CTP based temporal bi-directional search for the ALSP problem via a novel impromptu rendezvous termination condition. Theoretical and experimental analysis show that the proposed approach outperforms the related work approaches particularly when there are few critical-time-points.

Capacity-Constrained network-voronoi diagram: a summary of results

Given a graph and a set of service centers, a Capacity Constrained Network-Voronoi Diagram (CCNVD) partitions the graph into a set of contiguous service areas that meet service center capacities and minimize the sum of the distances (min-sum) from graph-nodes to allotted service centers. The CCNVD problem is important for critical societal applications such as assigning evacuees to shelters and assigning patients to hospitals. This problem is NP-hard; it is computationally challenging because of the large size of the transportation network and the constraint that Service Areas (SAs) must be contiguous in the graph to simplify communication of allotments. Previous work has focused on honoring either service center capacity constraints (e.g., min-cost flow) or service area contiguity (e.g., Network Voronoi Diagrams), but not both. We propose a novel Pressure Equalizer (PE) approach for CCNVD to meet the capacity constraints of service centers while maintaining the contiguity of service areas. Experiments and a case study using post-hurricane Sandy scenarios demonstrate that the proposed algorithm has comparable solution quality to min-cost flow in terms of min-sum; furthermore it creates contiguous service areas, and significantly reduces computational cost.

A Dartboard Network Cut Based Approach to Evacuation Route Planning: A Summary of Results

Given a transportation network, a population, and a set of destinations, the goal of evacuation route planning is to produce routes that minimize the evacuation time for the population. Evacuation planning is essential for ensuring public safety in the wake of man-made or natural disasters (e.g., terrorist acts, hurricanes, and nuclear accidents). The problem is challenging because of the large size of network data, the large number of evacuees, and the need to account for capacity constraints in the road network. Promising methods that incorporate capacity constraints into route planning have been developed but new insights are needed to reduce the high computational costs incurred by these methods with large-scale networks. In this paper, we propose a novel scalable approach that explicitly exploits the spatial structure of road networks to minimize the computational time. Our new approach accelerates the routing algorithm by partitioning the network using dartboard network-cuts and groups node-independent shortest routes to reduce the number of search iterations. Experimental results using a Minneapolis, MN road network demonstrate that the proposed approach outperforms prior work for CCRP computation by orders of magnitude.

Lagrangian Approaches to Storage of Spatio-Temporal Network Datasets

Given a spatio-temporal network (STN) and a set of STN operations, the goal of the Storing Spatio-Temporal Networks (SSTN) problem is to produce an efficient method of storing STN data that minimizes disk I/O costs for given STN operations. The SSTN problem is important for many societal applications, such as surface and air transportation management systems. The problem is NP hard, and is challenging due to an inherently large data volume and novel semantics (e.g., Lagrangian reference frame). Related works rely on orthogonal partitioning approaches (e.g., snapshot and longitudinal) and incur excessive I/O costs when performing common STN queries. Our preliminary work proposed a non-orthogonal partitioning approach in which we optimized the LGetOneSuccessor() operation that retrieves a single successor for a given node on STN. In this paper, we provide a method to optimize the LGetAllSuccessors() operation, which retrieves all successors for a given node on a STN. This new approach uses the concept of a Lagrangian Family Set (LFS) to model data access patterns for STN queries. Experimental results using real-world road and flight traffic datasets demonstrate that the proposed approach outperforms prior work for LGetAllSuccessors() computation workloads.

Spatio-temporal Networks: Modeling, Storing, and Querying Temporally-Detailed Roadmaps

Given spatio-temporal networks (e.g., roadmaps with traffic speed reported as a time-series in 5 min increments over a typical day for each road-segment) and operators (e.g., network snapshot, shortest path or path evaluation), a spatio-temporal network model provides a computer representation to facilitate reasoning, analysis and algorithm design for important societal applications. For example, next generation routing services are estimated to save consumers hundreds of billions of dollars in terms of time and fuel saved by 2020. Developing a model for spatio-temporal networks is challenging due to potentially conflicting requirements of expressiveness and model simplicity. Related work in Time Geography models spatio-temporal movement and relationships via dimension-based representations such as space-time prisms and space-time trajectories. These representations are not adequate for many STN use-cases, such as spatio-temporal routing queries. To address these limitations, we discuss a novel model called time-aggregated graph (TAG) that allows the properties of the network to be modeled as a time series. This model retains spatial network information while reducing the temporal replication needed in other models, thus resulting in a much more efficient model for several computational techniques for routing problems. In this chapter, we discuss spatio-temporal networks as represented by time-aggregated graphs at a conceptual, logical, and physical level. This chapter also focuses on shortest path algorithms for spatio-temporal networks. We develop the topics via case studies using TAGs in context of Lagrangian shortest-path queries and evacuation route planning.