Collaborating Authors

Planning & Scheduling

Why AI-optimized workflows aren't always best for business


Check out all the on-demand sessions from the Intelligent Security Summit here. Workflow and process inefficiencies can cost up to 40% of a company's annual revenue. In many instances, companies seek to resolve this issue by implementing Artificial Intelligence (AI) scheduling algorithms. This is seen as a beneficial tool for business models that depend on speed and efficiency, such as delivery services and the logistics sector. While AI has certainly helped with some of the time-consuming and often unpredictable tasks associated with scheduling workers across departments, the model is not yet perfect.

Automated Dynamic Algorithm Configuration

Journal of Artificial Intelligence Research

The performance of an algorithm often critically depends on its parameter configuration. While a variety of automated algorithm configuration methods have been proposed to relieve users from the tedious and error-prone task of manually tuning parameters, there is still a lot of untapped potential as the learned configuration is static, i.e., parameter settings remain fixed throughout the run. However, it has been shown that some algorithm parameters are best adjusted dynamically during execution. Thus far, this is most commonly achieved through hand-crafted heuristics. A promising recent alternative is to automatically learn such dynamic parameter adaptation policies from data. In this article, we give the first comprehensive account of this new field of automated dynamic algorithm configuration (DAC), present a series of recent advances, and provide a solid foundation for future research in this field. Specifically, we (i) situate DAC in the broader historical context of AI research; (ii) formalize DAC as a computational problem; (iii) identify the methods used in prior art to tackle this problem; and (iv) conduct empirical case studies for using DAC in evolutionary optimization, AI planning, and machine learning.

TOOLTANGO: Common sense Generalization in Predicting Sequential Tool Interactions for Robot Plan Synthesis

Journal of Artificial Intelligence Research

Robots assisting us in environments such as factories or homes must learn to make use of objects as tools to perform tasks, for instance, using a tray to carry objects. We consider the problem of learning common sense knowledge of when a tool may be useful and how its use may be composed with other tools to accomplish a high-level task instructed by a human. Specifically, we introduce a novel neural model, termed TOOLTANGO, that first predicts the next tool to be used, and then uses this information to predict the next action. We show that this joint model can inform learning of a fine-grained policy enabling the robot to use a particular tool in sequence and adds a significant value in making the model more accurate. TOOLTANGO encodes the world state, comprising objects and symbolic relationships between them, using a graph neural network and is trained using demonstrations from human teachers instructing a virtual robot in a physics simulator. The model learns to attend over the scene using knowledge of the goal and the action history, finally decoding the symbolic action to execute. Crucially, we address generalization to unseen environments where some known tools are missing, but unseen alternative tools are present. We show that by augmenting the representation of the environment with pre-trained embeddings derived from a knowledge-base, the model can generalize effectively to novel environments. Experimental results show at least 48.8-58.1% absolute improvement over the baselines in predicting successful symbolic plans for a simulated mobile manipulator in novel environments with unseen objects. This work takes a step in the direction of enabling robots to rapidly synthesize robust plans for complex tasks, particularly in novel settings.

Travel Smarter with AI and ML. Price prediction and comparison


One of the main ways that AI and ML can be used for travel hacking is by predicting and comparing prices for flights, hotels, and other travel accommodations. By analyzing past data on prices and demand, machine learning algorithms can predict the likelihood of price fluctuations and suggest the best time to book a trip. This can save travelers money and help them plan their trips more efficiently. AI and ML can also be used to make personalized recommendations for travel destinations and experiences based on an individual's preferences and past travel history. By analyzing a traveler's interests and past travel patterns, AI algorithms can suggest destinations and activities that are tailored to their interests and needs.

How quantum computing can navigate robots through crowded places


Movie scenes with robots and vehicles autonomously moving around and interacting with their environment are no longer science fiction but slowly becoming part of our daily lives. On Mars, robotic vehicles already travel around freely and perform tasks. However, on Earth, the setup is quite different: robots have to navigate through crowded places and plan their route numerous times per second to safely avoid pedestrians and property. Currently, the development of rapidly responsive robot path planning is a hot topic of research. In computational methods, a robot's ability to pivot heavily depends on continuously solving differential equations.

Human-instructed Deep Hierarchical Generative Learning for Automated Urban Planning


The essential task of urban planning is to generate the optimal land-use configuration of a target area. However, traditional urban planning is time-consuming and labor-intensive. Deep generative learning gives us hope that we can automate this planning process and come up with the ideal urban plans. While remarkable achievements have been obtained, they have exhibited limitations in lacking awareness of: 1) the hierarchical dependencies between functional zones and spatial grids; 2) the peer dependencies among functional zones; and 3) human regulations to ensure the usability of generated configurations. To address these limitations, we develop a novel human-instructed deep hierarchical generative model.

Asymmetric Action Abstractions for Planning in Real-Time Strategy Games

Journal of Artificial Intelligence Research

Action abstractions restrict the number of legal actions available for real-time planning in zero-sum extensive-form games, thus allowing algorithms to focus their search on a set of promising actions. Even though unabstracted game trees can lead to optimal policies, due to real-time constraints and the tree size, they are not a practical choice. In this context, we introduce an action abstraction scheme which we call asymmetric action abstraction. Asymmetric abstractions allow search algorithms to "pay more attention" to some aspects of the game by unevenly dividing the algorithm's search effort amongst different aspects of the game. We also introduce four algorithms that search in asymmetrically abstracted game trees to evaluate the effectiveness of our abstraction schemes. Two of our algorithms are adaptations of algorithms developed for searching in action-abstracted spaces, Portfolio Greedy Search and Stratified Strategy Selection, and the other two are adaptations of an algorithm developed for searching in unabstracted spaces, NaïveMCTS. An extensive set of experiments in a real-time strategy game shows that search algorithms using asymmetric abstractions are able to outperform all other search algorithms tested.

AI in Supply Chain and Logistics


It is very easy to organize the various requirements and deadlines using traditional planning techniques, such as material requirements planning and its derivatives. They create a plan or schedule that seems logical and might actually function by applying a fixed logic to the vast amount of data. A typical computer-generated plan, however, may include elements that are unrealistic or outright impossible, such as having a work center commit to three or four times the amount of production that it can complete in a day. This is something that any experienced planner will tell you. The latest generation of so-called advanced planning and scheduling (APS) software, which trades off priorities and alternatives instead of using fixed logic to build a more workable plan that balances resources and material dates and amounts, overcomes many of these restrictions.