Despite an increasing number of successful interventions designed to broaden participation in computing research, there is still significant attrition among historically marginalized groups in the computing research pipeline. This experience report describes a first-of-its-kind Undergraduate Consortium (UC; https://aaai-uc.github.io/about) that addresses this challenge by empowering students with a culmination of their undergraduate research in a conference setting. The UC, conducted at the AAAI Conference on Artificial Intelligence (AAAI), aims to broaden participation in the AI research community by recruiting students, particularly those from historically marginalized groups, supporting them with mentorship, advising, and networking as an accelerator toward graduate school, AI research, and their scientific identity. This paper presents our program design, inspired by a rich set of evidence-based practices, and a preliminary evaluation of the first years that points to the UC achieving many of its desired outcomes. We conclude by discussing insights to improve our program and expand to other computing communities.

The controllability of a temporal network is defined as an agent's ability to navigate around the uncertainty in its schedule and is well-studied for certain networks of temporal constraints. However, many interesting real-world problems can be better represented as Probabilistic Simple Temporal Networks (PSTNs) in which the uncertain durations are represented using potentially-unbounded probability density functions. This can make it inherently impossible to control for all eventualities. In this paper, we propose two new dynamic controllability algorithms that attempt to maximize the likelihood of successfully executing a schedule within a PSTN. The first approach, which we call Min-Loss DC, finds a dynamic scheduling strategy that minimizes loss of control by using a conflict-directed search to decide where to sacrifice the control in a way that optimizes overall success. The second approach, which we call Max-Gain DC, works in the other direction: it finds a dynamically controllable schedule and then attempts to progressively strengthen it by capturing additional uncertainty. Our approaches are the first known that work by finding maximally dynamically controllable schedules. We empirically compare our approaches against two existing PSTN offline dispatch approaches and one online approach and show that our Min-Loss DC algorithm outperforms the others in terms of maximizing execution success while maintaining competitive runtimes.

Increased space and ground sensing is enabling dramatic new measurements of a wide range of Earth Science and Applied Earth Science phenomena, including: volcanism, flooding, wildfires, weather, and many other phenomena. New challenges exist to rapidly assimilate available data and to optimize measurements (e.g. direct assets) to best observe these complex and dynamic spatiotemporal phenomena. Artificial Intelligence offers the potential to assist in data interpretation and resource allocation to best allocate sensing assets. We describe efforts to build and experiment with such “sensorweb” systems and offer some direction for the future sensorweb observation systems.

Controllability for Simple Temporal Networks with Uncertainty (STNUs) has thus far been limited to three levels: strong, dynamic, and weak. Because of this, there is currently no systematic way for an agent to assess just how far from being controllable an uncontrollable STNU is. We provide new insights inspired by a geometric interpretation of STNUs to introduce the degrees of strong and dynamic controllability — continuous metrics that measure how far a network is from being controllable. We utilize these metrics to approximate the probabilities that an STNU can be dispatched successfully offline and online respectively. We introduce new methods for predicting the degrees of strong and dynamic controllability for uncontrollable networks. We further generalize these metrics by defining likelihood of controllability, a controllability measure that applies to Probabilistic Simple Temporal Networks (PSTNs). Finally, we empirically demonstrate that these metrics are good predictors of actual dispatch success rate for STNUs and PSTNs.

A critical challenge in temporal planning is robustly dealing with non-determinism, e.g., the durational uncertainty of a robot's activity due to slippage or other unexpected influences. Recent advances show that robustness is a better measure of solution quality than traditional metrics such as flexibility. This paper introduces the Robust Execution Problem for finding maximally robust dispatch strategies for general probabilistic temporal planning problems. While generally intractable, we introduce approximate solution techniques — one that can be computed statically prior to the start of execution with robustness guarantees and one that dynamically adjusts to opportunities and setbacks during execution. We show empirically that our dynamic approach outperforms all known approaches in terms of execution success rate.

Trust and cooperation are fundamental to human interactions. How much we trust other people directly influences the decisions we make and our willingness to cooperate. It thus seems natural that trust be equally important in successful human-robot interaction (HRI), since how much a human trusts a robot affects how they might interact with it. We propose using a coin entrustment game, a variant of prisoner’s dilemma, to measure trust and cooperation as separate phenomenon between human and robot agents. With this game, we test the following hypotheses: (1) Humans will achieve and maintain higher levels of trust when interacting with what they believe to be a robot than with another human; and (2) humans will cooperate more readily with robots and will maintain a higher level of cooperation. This work contributes an experimental paradigm that uses the coin entrustment game as a way to test our hypotheses. Our empirical analysis shows that humans tend to trust robots to a greater degree than other humans, while cooperating equally well in both.

Flexibility in agent scheduling increases the resilience of temporal plans in the face of new constraints. However,current metrics of flexibility ignore domain knowledge about how such constraints might arise in practice, e.g., due to the uncertain duration of a robot’s transitiontime from one location to another. Probabilistic temporalplanning accounts for actions whose uncertain durations can be modeled with probability density functions. We introduce a new metric called robustness that measures the likelihood of success for probabilistic temporalplans. We show empirically that in multi-robot planning,robustness may be a better metric for assessing the quality of temporal plans than flexibility, thus reframing many popular scheduling optimization problems.

College students often struggle to balance their work with personal wellness. In part, this occurs because students work when they are unable to focus. We hypothesize that we can adapt the Experience Sampling Method (ESM) to build a model of users’ efficacy and predict when they will be most likely to experience flow, a state of motivation and immersion. We also hypothesize that we can present this information effectively to users, allowing them to understand when they are most likely to achieve flow. In order to test these hypotheses, we introduce the Productivity and Wellness Pal (PaWPal), a smartphone-based application that seeks to make users aware of their efficacy at various tasks as well as which courses of action are likely to lead to immersive experiences.

There exists an increasing demand to incorporate mobile interactive robots to assist humans in repetitive, non-value added tasks in the manufacturing domain. Our aim is to develop a mobile robotic assistant for fetch-and-deliver tasks in human-oriented assembly line environments. Assembly lines present a niche yet novel challenge for mobile robots; the robot must precisely control its position on a surface which may be either stationary, moving, or split (e.g. in the case that the robot straddles the moving assembly line and remains partially on the stationary surface). In this paper we present a control and sensing solution for a mobile robotic assistant as it traverses a moving-floor assembly line. Solutions readily exist for control of wheeled mobile robots on static surfaces; we build on the open-source Robot Operating System (ROS) software architecture and generalize the algorithms for the moving line environment. Off-the-shelf sensors and localization algorithms are explored to sense the moving surface, and a customized solution is presented using PX4Flow optic flow sensors and a laser scanner-based localization algorithm. Validation of the control and sensing system is carried out both in simulation and in hardware experiments on a customized treadmill. Initial demonstrations of the hardware system yield promising results; the robot successfully maintains its position while on, and while straddling, the moving line.

The Multiagent Disjunctive Temporal Problem (MaDTP) is a general constraint-based formulation for scheduling problems that involve interdependent agents. Decoupling agents' interdependent scheduling problems, so that each agent can manage its schedule independently, requires agents to adopt additional local constraints that effectively subsume their interdependencies. In this paper, we present the first algorithm for decoupling MaDTPs. Our distributed algorithm is provably sound and complete. Our experiments show that the relative efficiency of using temporal decoupling to find solution spaces for MaDTPs, compared to algorithms that find complete solution spaces, improves with the interconnectedness between agents schedules, leading to orders of magnitude relative speeedup. However, decoupling by its nature restricts agents' scheduling flexibility; we define novel flexibility metrics for MaDTPs, and show empirically how the flexibility sacrificed depends on the degree of coupling between agents' schedules.

This research focuses on building foundational algorithms for scheduling agents that assist people in managing their activities in environments where tempo and complex activity interdependencies outstrip people's cognitive capacity. We address the critical challenge of reasoning over individuals' interacting schedules to efficiently answer queries about how to meet scheduling goals while respecting individual privacy and autonomy to the extent possible. We formally define the Multiagent Simple Temporal Problem for naturally capturing and reasoning over the distributed but interconnected scheduling problems of multiple individuals. Our hypothesis is that combining bottom-up and top-down approaches will lead to effective solution techniques. In our bottom-up phase, an agent externalizes constraints that compactly summarize how its local subproblem affects other agents' subproblems, whereas in our top-down phase an agent proactively constructs and internalizes new local constraints that decouple its subproblem from others'. We confirm this hypothesis by devising distributed algorithms that calculate summaries of the joint solution space for multiagent scheduling problems, without centralizing or otherwise redistributing the problems. The distributed algorithms permit concurrent execution to achieve significant speedup over the current art and also increase the level of privacy and independence in individual agent reasoning. These algorithms are most advantageous for problems where interactions between the agents are sparse compared to the complexity of agents' individual problems.

We introduce the Multiagent Disjunctive Temporal Problem (MaDTP), a new distributed formulation of the widely-adopted Disjunctive Temporal Problem (DTP) representation. An agent that generates a summary of all viable schedules, rather than a single schedule, can be more useful in dynamic environments. We show how a (Ma)DTP with the properties of minimality and decomposability provides a particularly efficacious solution space summary.However, in the multiagent case, these properties sacrifice an agent's strategic interests while incurring significant computational overhead. We introduce a new property called local decomposability that exploits loose-coupling between agents' problems, protects strategic interests, and supports typical queries. We provide and evaluate a new distributed algorithm that summarizes agents' solution spaces in significantly less time and space by using local, rather than full, decomposability.

Consider Ann’s morning scheduling problem. Ann is a graduate student, who, among many other objectives, would like to both exercise and work on research during her morning. I summarize possible morning activities in Table 1. Even for these two simple objectives, selecting a feasible schedule from the many possible schedules (run then generate experimental results, write then bike, swim then read some related work, etc.) may be non-trivial. However, suppose Ann wants to run with her friend Bill, who notoriously oversleeps his alarm. Additionally, suppose also that Ann must coordinate the use of her lab’s computational cluster with her lab mate Claire. Without further information from Bill and Claire, it is impossible for Ann to determine which candidate schedules will successfully achieve her morning goals. One option for Ann would be to myopically select her schedule anyway, with the risk that her attempt to run with Bill or to use the computational cluster could result in a failed goal. As another option, Ann could also volunteer to collect the scheduling constraints of both Bill and Claire and generate a single joint morning schedule. However, this puts additional scheduling burden on Ann while requiring both Bill and Claire to reveal other scheduling commitments they may prefer to keep private. Even if Ann employed a centralized computational agent to solve this global scheduling problem, the resulting combinatorics may limit the scalability of such a centralized approach. Instead, the pervasiveness of personal computational devices, coupled with desires for scalability and privacy, argue for decentrally solving such problems using multiagent algorithms. My thesis focuses on providing scalable, multiagent algorithms for solving rich, complex multiagent activity scheduling and selection problems, while retaining as much privacy as possible on behalf of the human users. My approach is distinct from other recent multiagent scheduling approaches (Hunsberger 2002; Smith et al. 2007; Shah, Conrad, and Williams 2009) in that it uses a constraint-based representation of selection (finite-domain) aspects of scheduling problems in addition to the scheduling aspects. I proceed by introducing the Multiagent Selection and Scheduling Problem