Eric Beaudry

Spartacus attending the 2005 AAAI conference

Spartacus is our robot entry in the 2005 AAAI Mobile Robot Challenge, making a robot attend the National Conference on Artificial Intelligence. Designing robots that are capable of interacting with humans in real-life settings can be considered the ultimate challenge when it comes to intelligent autonomous systems. One key issue is the integration of multiple modalities (e.g., mobility, physical structure, navigation, vision, audition, dialogue, reasoning). Such integration increases the diversity and also the complexity of interactions the robot can generate. It also makes it difficult to monitor how such increased capabilities are used in unconstrained conditions, whether it is done while the robot is in operation of afterwards. This paper reports solutions and findings resulting from our hardware, software and decisional integration work on Spartacus. It also outlines perspectives in making intelligent and interaction capabilities evolve for autonomous robots.

Modularity and integration in the design of a socially interactive robot

Designing robots that are capable of interacting with humans in real life settings is a challenging task. One key issue is the integration of multiple modalities (e.g., mobility, physical structure, navigation, vision, audition, dialogue, reasoning) into a coherent framework. Taking the AAAI mobile robot challenge (making a robot attend the national conference on artificial intelligence) as the experimental context, we are currently addressing hardware, software and computation integration issues involved in designing a robot capable of sophisticated interaction with humans. This paper reports on our design solutions and the current status of the work, along with the potential impacts this design on human-robot interaction research.

Detecting information under and from shadow in panchromatic Ikonos images of the city of Sherbrooke

The presence of shadow is increasingly alarming on the images with very high spatial resolution mainly in urban area. After shadow detection, how to give an added value to detected shadow? Some kinds of exploitation are presented: the restitution of information under shadow and the deduction of object height from their shadow. The restitution of surfaces under shadow is based on the analysis of contextual and textural information between the shadow and its neighbouring surfaces not in sun side. Assuming that the same surface texture is independent of shadow. The height of objects is deduced knowing the length of their projected shadow on the ground, the position (azimuth and zenith) of the sun and the sensor at the time of acquisition. These exploitations of shadow were carried out on a panchromatic Ikonos image of Sherbrooke. Results are validated with land use map for surfaces under shadow, and measured height for the building height from shadow

Motion planning for an omnidirectional robot with steering constraints

Omnidirectional mobile robots, i.e., robots that can move in any direction without changing their orientation, offer better manoeuvrability in natural environments. Modeling the kinematics of such robots is a challenging problem and different approaches have been investigated. One of the best approaches for a nonholonomic robot is to model the robots velocity state as the motion around its instantaneous center of rotation (ICR). In this paper, we present a motion planner designed to compute efficient trajectories for such a robot in an environment with obstacles. The action space is modeled in terms of changes of the ICR and the motion around it. Our motion planner is based on a Rapidly-Exploring Random Trees (RRT) algorithm to sample the action space and find a feasible trajectory from an initial configuration to a goal configuration. To generate fluid paths, we introduce an adaptive sampling technique taking into account constraints related to the ICR-based action space.

Planning for a mobile robot to attend a conference

The AAAI Mobile Robot Challenge requires robots to start from the entrance of the conference site, find their own way to the registration desk, socially interact with people and perform volunteer duties as required, then report at a prescribed time in a conference hall to give a talk and answer questions These specifications convey some interesting planning problems that appear to be too complex for some of the most efficient AI planning systems that we analyzed Based on this analysis, we present a new planning approach that we are developing to meet the challenge Preliminary results show that our approach performs much better on robot conference planning problems than any of the other AI planning systems we tested.

Using Markov decision theory to provide a fair challenge in a roll-and-move board game

Board games are often taken as examples to teach decision-making algorithms in artificial intelligence (AI). These algorithms are generally presented with a strong focus on winning the game. Unfortunately, a few important aspects, such as the gaming experience of human players, are often missing from the equation. This paper presents a simple board game we use in an introductory course in AI to initiate students to the gaming experience issue. The Snakes and Ladders game has been modified to provide different levels of challenges for students. The game with such modifications offers theoretical, algorithmic and programming challenges. One of the most complex is the generation of an optimal policy to provide a fair challenge to an opponent. A solution based on Markov Decision Processes (MDPs) is presented. This approach relies on a simple model of the opponents playing behaviour.

CORALS: A real-time planner for anti-air defense operations

Forces involved in modern conflicts may be exposed to a variety of threats, including coordinated raids of advanced ballistic and cruise missiles. To respond to these, a defending force will rely on a set of combat resources. Determining an efficient allocation and coordinated use of these resources, particularly in the case of multiple simultaneous attacks, is a very complex decision-making process in which a huge amount of data must be dealt with under uncertainty and time pressure. This article presents CORALS (COmbat Resource ALlocation Support), a real-time planner developed to support the command team of a naval force defending against multiple simultaneous threats. In response to such multiple threats, CORALS uses a local planner to generate a set of local plans, one for each threat considered apart, and then combines and coordinates them into a single optimized, conflict-free global plan. The coordination is performed through an iterative process of plan merging and conflict detection and resolution, which acts as a plan repair mechanism. Such an incremental plan repair approach also allows adapting previously generated plans to account for dynamic changes in the tactical situation.

Reactive planning as a motivational source in a behavior-based architecture

Behavior-based architectures use behaviors as building blocks for decision-making and action execution processes. Behaviors are distributed and evaluated in parallel for the control of the robot, taking real-time inputs from sensory data and sending real-time commands to effectors. No centralized components exist in these architectures, each module carrying out its own strategy independently, making an overall behavior emerge from the interaction between the concurrently executed modules and the environment. In this paper, we discuss the use of a reactive hierarchical task network (HTN) planner in a behavior-based robot architecture. The planner in this architecture is not a central component on which everything else relies on, but acts as one of the motivational modules recommending tasks to be executed and influencing the selection and configuration of behaviors. The planning module allows the behavior-based architecture to deal with tasks with priorities, flexible time constraints and on-line planning using a simple but very effective reactive planning strategy. We demonstrate our approach in the context of making a robot attend a conference.

Using partial satisfaction planning to automatically select NPCs' goals and generate plans in a simulation game

Controlling and coordinating a large number of Non-Playable Characters (NPCs) is an important challenge in video games. In order to obtain a realistic behaviour, traditional approaches relies hand-written rule based scripts or finite state machines. In the last decade, a new approach to artificial intelligence has emerged. Indeed, automated planning is now a way to control NPCs. However, the gap between theory and application is still quite not filled. In this paper, we address different approaches of i) goal selection and distribution for autonomous planning agents by a coordination module and ii) autonomous goal selection by planning agents using partial satisfaction planning. Our techniques result in a simpler way to coordinate NPCs, still effective in terms of CPU and produce realistic behaviour.