Technology Overview

This research area focuses on the use of analytics to improve the day-to-day operations as well as manage the experience of visitors of large leisure-oriented facilities such as theme parks and expos. Key research efforts within this area include (a) the use of decision analytics and agent-based technologies to optimize the visit sequences of thousands of visitors to theme parks under dynamic changes in the status of park attractions and (b) the use of statistical optimization techniques to devise revenue-maximizing bundled packages for visitors.

The key projects in this research area include:

1. Optimized Crowd Guidance under Dynamic Operating Environments
2. Spatiotemporal Machine Learning & Queuing Analytics
3. Optimization of Bundle Design

1. Orienteering Problem (OP). This is the anchor problem behind route guidance. It is a variant of the well-known prize-collecting traveling salesman problem, where the salesman needs to choose a subset of cities to visit within a given deadline. We worked on methods for several extensions of OP, including Time-Dependent OP, Team OP, Dynamic OP, Stochastic OP, Dynamic Stochastic OP, Multi-Agent OP, and Selfish OP.
   
   Visit our Orienteering Problem Library (OPLib) page here.


2. Incentives as a mechanism for crowd coordination (or behavioural influence). Congestion occurs when there is competition for resources by selfish agents. In this work, we developed an optimization model for smoothing out congestion in a network of resources by using personalized well-timed incentives. Our model computes an equilibrium solution for the resource sharing congestion game with incentives and budget constraints.


3. Agent-based simulator. Using the data collected from our ground surveys, we quantified visitor behaviors and eventually incorporated them into a massive agent-based simulation where up to 15,000 visitor agents are modeled. We demonstrate how the simulator can be used to understand the crowd build-up and the impacts of various control policies (such as route guidance and incentives) on visitor behavior.

We work on a challenging data-driven dynamic queue prediction algorithm on a network of queues, based on a dataset provided by a theme park operator on customers’ trajectories. We developed a model that combine expectation-maximization (with multiple random variables to capture user behaviour between attractions) estimation with regression analysis to isolate and estimate individual queue delays in a mix of heterogeneous attractions. The statistical inferences capture variations among different categories of ticket holders in terms of their cross-attraction movement dynamics and arrival patterns and also help identify the clustering (arrival and movement in groups) behaviour that occurs across different users. These inferences were then fed into our agent-based simulator to help predict the queue dynamics and user behaviour that may be expected at future events.

We work on the analysis of the visitor trajectory patterns of actual visits of attractions in a theme park, as well as the corresponding sales data. We worked on the analysis of these datasets and developed the first-of-its-kind data-driven revenue management model based on the dataset that provides decision support on bundle design. In technical terms, we provided an extension to the well-known Markov Random Fields representation to capture interactions between the attractions and hence the attractiveness of various bundles. We also designed scalable algorithms that exploit the notions of independence and correlation in designing bundles.

• DIRECT: A Scalable Approach for Route Guidance in Selfish Orienteering Problems. Pradeep Varakantham, Hala Mostafa, Na Fu and Hoong Chuin Lau, In Proceedings of 14th International Conference on Autonomous Agents and Multiagent Systems (AAMAS’15), Istanbul, Turkey, May 2015.


• A Mathematical Model and Metaheuristics for Time Dependent Orienteering Problem. Aldy Gunawan, Hoong Chuin Lau and Zhi Yuan, 10th International Conference on the Practice and Theory of Automated Timetabling (PATAT’14), York, United Kingdom, August 2014.


• Multi-Agent Orienteering Problem. C. Chen, S. F. Cheng and H. C. Lau, Web Intelligence and Agent Systems Journal.


• Budgeted Personalized Incentive Approaches for Smoothing Out Congestion in Resource Networks. Pradeep Varakantham, Na Fu, William Yeoh, Shih-Fen Cheng and Hoong Chuin Lau, In Proceedings of Algorithmic Decision Theory (ADT) (Lecture Notes in Computer Science), November 2013.


• Anonymous Authentication of Visitors for Mobile Crowd Sensing at Amusement Parks. Divyan Munirathnam KONIDALA, Robert H. DENG, Yingjiu LI, Hoong Chuin LAU and Stephen E. FIENBERG, 9th International Conference on Information Security Practice and Experience (ISPEC’13), Lanzhou, China, May 2013.


• Lagrangian Relaxation for Large-Scale Multi-Agent Planning. Geoffrey J. GORDON, Pradeep VARAKANTHAM, William YEOH, Hoong Chuin LAU, Ajay S. ARAVAMUDHAN and Shih-Fen CHENG, 12th IEEE/WIC/ACM International Conference on Intelligent Agent Technology (IAT’12), Macau, China, December 2012.


• Dynamic Stochastic Orienteering Problems for Risk-Aware Applications. Hoong Chuin Lau, William Yeoh, Pradeep Varakanham, Duc Thien Nguyen and Huaxing Chen, In Proceedings of Uncertainty in AI (UAI), August 2012.


• Lagrangian Relaxation for Large-Scale Multi-Agent Planning. Geoffrey J. GORDON, Pradeep VARAKANTHAM, William YEOH, Hoong Chuin LAU, Ajay S. ARAVAMUDHAN and Shih-Fen CHENG, 11th International Conference on Autonomous Agents and Multiagent Systems (AAMAS’12), Valencia, Spain, Jun 2012.


• Theme Park Navigation Application. Hoong Chuin LAU, Chun Pong Fan and Elizabeth LIM, Source:Website of National Research Foundation under the Singapore Technologies & Innovations Showcase – Interactive Digital Media, August 2014.


• New Applications Developed to Enhance Theme Park Experience. Hoong Chuin LAU, Chun Pong Fan and Elizabeth LIM, Source: Lianhe Zaobao, March 2014.

• Coordinating Mobile Crowdworkers with Stochastic Routine Routes. Chen Cen, Shih Fen Cheng, Hoong Chuin Lau and Archan Misra, Extended abstract in Proceedings at the 14th International Conference on Autonomous Agents and Multiagent Systems (AAMAS’15), Istanbul, Turkey, May 2015.


• On Understanding Diffusion Dynamics of Patrons at a Theme Park.Jiali DU, Akshat Kumar and Pradeep Varakantham, Extended abstract at the 13th International Conference on Autonomous Agents and Multiagent Systems (AAMAS’14), Paris, France, May 2014.


• A Quantitative Analysis of Decision Process in Social Groups Using Human Trajectories. Truc Viet LE, Siyuan LIU, Hoong Chuin LAU and Ramayya KRISHNAN, 13th International Conference on Autonomous Agents and Multi-Agent Systems (AAMAS’14), Paris, France, May 2014.


• “Network-Theoretic” Queuing Delay Estimation in Theme Park Attractions. H Ajay Aravamudhan and Archan Misra and Hoong Chuin Lau, In Proceedings of 9th IEEE International Conference on Automation Science and Engineering (IEEE CASE’13), Madison, Wisconsin, USA, August 2013.


• An Agent-Based Simulation Approach to Experience Management in Theme Parks. S. F. Cheng, L. Lin, J. Du, H. C. Lau and P. Varakantham, In Proceedings of Winter Simulation Conference 2013 (WCS’13), Washington DC, USA, December 2013.

• Predicting Bundles of Spatial Locations from Learning Revealed Preference Data. Truc Viet Le, Siyuan Liu, Hoong Chuin Lau and Ramayya Krishnan, In Proceedings of 14th International Conference on Autonomous Agents and Multiagent Systems (AAMAS’15), Istanbul, Turkey, May 2015.


• Interacting Knapsack Problem in Designing Resource Bundles. H. Nguyen, P. Varankantham, H. C. Lau and S. F. Cheng, In Proceedings of 10th Metaheuristics International Conference (MIC’13), Singapore, Augest 2013.