Mikhail Volkov

ChangiNOW: A mobile application for efficient taxi allocation at airports

We present an application that uses a predictive queueing model to efficiently allocate taxis. The system uses observed taxi and flight data at each of the four terminals of Singapores Changi Airport to estimate the expected waiting time and queue length for taxis arriving at these terminals, and then sends taxis to terminals where demand is highest. We propose a service model that enables our system to be deployed on a smartphone platform to participating taxi drivers. We present the theoretical details which underpin our prediction engine and corroborate our theory with several targeted numerical simulations. Finally, we evaluate the performance of this system in large-scale experiments and show that our system achieves a significant improvement in both passenger and taxi waiting time.

Coresets for visual summarization with applications to loop closure

In continuously operating robotic systems, efficient representation of the previously seen camera feed is crucial. Using a highly efficient compression coreset method, we formulate a new method for hierarchical retrieval of frames from large video streams collected online by a moving robot. We demonstrate how to utilize the resulting structure for efficient loop-closure by a novel sampling approach that is adaptive to the structure of the video. The same structure also allows us to create a highly-effective search tool for large-scale videos, which we demonstrate in this paper. We show the efficiency of proposed approaches for retrieval and loop closure on standard datasets, and on a large-scale video from a mobile camera.

Markov-based redistribution policy model for future urban mobility networks

In this paper we present a Markov-based urban transportation model that captures the operation of a fleet of taxis in response to incident customer arrivals throughout the city. We consider three different evaluation criteria: (1) minimizing the number of transportation resources for urban planning; (2) minimizing fuel consumption for the drivers; and (3) minimizing customer waiting time increase the overall quality of service. We present a practical policy and evaluate it by comparing against the actual observed redistribution of taxi drivers in Singapore. We show through simulation that our proposed policy is stable and improves substantially upon the default unmanaged redistribution of taxi drivers in Singapore with respect to the three evaluation criteria.

Dynamic Patrolling Policy for Optimizing Urban Mobility Networks

A dynamic patrolling policy is presented for a fleet of service vehicles operating in response to incident requests in an urban transportation network. We modify an existing adaptive, informative path controller so that the fleet of vehicles is driven to locally optimal service configurations within the environment. These configurations, called patrolling loops, minimize the distance between the instantaneous vehicle position and incident customer request. Our patrolling algorithm is trained using one month of data from a fleet of 16,000 vehicles. This historical dataset is used to learn the parameters required to set up a representative urban mobility model. Using this model we conduct large-scale simulations to show the global stability of the patrolling policy and evaluate the performance of our system by comparing it against a greedy service policy and historical data.

Machine learning and coresets for automated real-time video segmentation of laparoscopic and robot-assisted surgery

Context-aware segmentation of laparoscopic and robot assisted surgical video has been shown to improve performance and perioperative workflow efficiency, and can be used for education and time-critical consultation. Modern pressures on productivity preclude manual video analysis, and hospital policies and legacy infrastructure are often prohibitive of recording and storing large amounts of data. In this paper we present a system that automatically generates a video segmentation of laparoscopic and robot-assisted procedures according to their underlying surgical phases using minimal computational resources, and low amounts of training data. Our system uses an SVM and HMM in combination with an augmented feature space that captures the variability of these video streams without requiring analysis of the nonrigid and variable environment. By using the data reduction capabilities of online k-segment coreset algorithms we can efficiently produce results of approximately equal quality, in realtime. We evaluate our system in cross-validation experiments and propose a blueprint for piloting such a system in a real operating room environment with minimal risk factors.

Environment characterization for non-recontaminating frontier-based robotic exploration

This paper addresses the problem of obtaining a concise description of a physical environment for robotic exploration. We aim to determine the number of robots required to clear an environment using non-recontaminating exploration. We introduce the medial axis as a configuration space and derive a mathematical representation of a continuous environment that captures its underlying topology and geometry. We show that this representation provides a concise description of arbitrary environments, and that reasoning about points in this representation is equivalent to reasoning about robots in physical space. We leverage this to derive a lower bound on the number of required pursuers. We provide a transformation from this continuous representation into a symbolic representation. Finally, we present a generalized pursuit-evasion algorithm. Given an environment we can compute how many pursuers we need, and generate an optimal pursuit strategy that will guarantee the evaders are detected with the minimum number of pursuers.