It is claimed that, in the nascent ‘Cognitive Era’, intelligent systems will be trained using machine learning techniques rather than programmed by software developers. A contrary point of view argues that machine learning has limitations, and, taken in isolation, cannot form the basis of autonomous systems capable of intelligent behaviour in complex environments. In this paper, we explore the contributions that agent-oriented programming can make to the development of future intelligent systems. We briefly review the state of the art in agent programming, focussing particularly on BDI-based agent programming languages, and discuss previous work on integrating AI techniques (including machine learning) in agent-oriented programming. We argue that the unique strengths of BDI agent languages provide an ideal framework for integrating the wide range of AI capabilities necessary for progress towards the next-generation of intelligent systems. We identify a range of possible approaches to integrating AI into a BDI agent architecture. Some of these approaches, e.g., ‘AI as a service’, exploit immediate synergies between rapidly maturing AI techniques and agent programming, while others, e.g., ‘AI embedded into agents’ raise more fundamental research questions, and we sketch a programme of research directed towards identifying the most appropriate ways of integrating AI capabilities into agent programs.

We present an agent-based model of the key activities of a troop of chacma baboons ( Papio hamadryas ursinus ) based on the data collected at De Hoop Nature Reserve in South Africa. We analyse the predictions of the model in terms of how well it is able to duplicate the observed activity patterns of the animals and the relationship between the parameters that control the agent's decision procedure and the model's predictions. At the current stage of model development, we are able to show that across a wide range of decision parameter values, the baboons are able to achieve their energetic and social time requirements. The simulation results also show that decisions concerning movement (group action selection) have the greatest influence on the outcomes. Those cases where the model's predictions fail to agree with the observed activity patterns have highlighted key elements that were missing from the field data, and that would need to be collected in subsequent fieldwork. Based on our experience, we believe group decision making is a fertile field for future research, and agent-based modelling offers considerable scope for understanding group action selection.

Alternating Time Temporal Logic (ATL) is widely used for the verification of multi-agent systems. We consider Resource Agent Logic (RAL), which extends ATL to allow the verification of properties of systems where agents act under resource constraints. The model checking problem for RAL with unbounded production and consumption of resources is known to be undecidable. We review existing (un)decidability results for fragments of RAL, tighten some existing undecidability results, and identify several aspects which affect decidability of model checking. One of these aspects is the availability of a ‘do nothing’, or idle action, which does not produce or consume resources. Analysis of undecidability results allows us to identify a significant new fragment of RAL for which model checking is decidable

We propose a framework for modelling situated resource-bounded agents. The framework is based on an objective ascription of intentional modalities and can be easily tailored to the system we want to model and the properties we wish to specify. As an elaboration of the framework, we introduce a logic, OBA, for describing the observations, beliefs, goals and actions of simple agents, and show that OBA is complete, decidable and has an efficient model checking procedure, allowing properties of agents specified in OBA to be verified using standard theorem proving or model checking techniques.

Synthetic populations are representations of actual individuals living in a specific area. They play an increasingly important role in studying and modeling individuals and are often used to build agent-based social simulations. Traditional approaches for synthesizing populations use a detailed sample of the population (which may not be available) or combine data into a single joint distribution, and draw individuals or households from these. The latter group of existing sample-free methods fail to integrate (1) the best available data on spatial granular distributions, (2) multi-variable joint distributions, and (3) household level distributions. In this paper, we propose a sample-free approach where synthetic individuals and households directly represent the estimated joint distribution to which attributes are iteratively added, conditioned on previous attributes such that the relative frequencies within each joint group of attributes are maintained and fit granular spatial marginal distributions. In this paper we present our method and test it for the Zuid-West district of The Hague, the Netherlands, showing that spatial, multi-variable and household distributions are accurately reflected in the resulting synthetic population.