Unmanned Aerial Vehicles (UAVs) offer new ways to fulfil a variety of urban transportation and service solutions. The ability to successfully plan and re-plan paths across a complex urban environment remains an unsolved significant problem. New Q-learning approaches have potential to address this problem, however they must first learn complex environment spaces. A traditional challenge within this field is the selection of suitable learning hyper-parameters that assist a Q-learning algorithm in achieving an optimal learning policy. It is known that testing and evaluating multiple hyper-parameter combinations is computationally expensive. Thus, this paper proposes a new method for hyper-parameter self-tuning, cyclically assigning hyper-parameters within a single learning process, eliminating the need to experimentally seek optimal hyper-parameter value combinations. Evaluation of the captured results show, training with cyclical hyper-parameter exploration instead of fixed values, achieves improved path generation, while reducing the cumulative learning time required. Although the focus of this approach is centred around a Multi Q-table Path Planning solution, this work presents a practical tool applicable to Reinforcement Learning techniques generally.

This paper introduces an original multi destination path planning approach for Unmanned Aerial Vehicles (UAVs) named MQTPP (Multi Q-Table Path Planning). MQTPP aims to reduce the computational burden of cyclical/continuous path planning through a Q-learning planning process whilst overcoming the fixed path origin problem. The preliminary performance evaluation results indicate that MQTPP performs well for longer paths, and allows for more efficient re-planning should mission objectives or environmental topography change.

A self-adaptive system adjusts its configuration to tolerate changes in its operating environment. To date, requirements modeling methodologies for self-adaptive systems have necessitated analysis of all potential system configurations, and the circumstances under which each is to be adopted. We argue that, by explicitly capturing and modelling uncertainty in the operating environment, and by verifying and analysing this model at runtime, it is possible for a system to adapt to tolerate some conditions that were not fully considered at design time. We showcase in this paper our tools and research results.

In earlier work we proposed the idea of requirements-aware systems that could introspect about the extent to which their goals were being satisfied at runtime. When combined with requirements monitoring and self adaptive capabilities, requirements awareness should help optimize goal satisfaction even in the presence of changing run-time context. In this paper we describe initial progress towards the realization of requirements-aware systems with REAssuRE. REAssuRE focuses on explicit representation of assumptions made at design time. When such assumptions are shown not to hold, REAssuRE can trigger system adaptations to alternative goal realization strategies.

Requirements awareness should help optimize requirements satisfaction when factors that were uncertain at design time are resolved at runtime. We use the notion of claims to model assumptions that cannot be verified with confidence at design time. By monitoring claims at runtime, their veracity can be tested. If falsified, the effect of claim negation can be propagated to the system's goal model and an alternative means of goal realization selected automatically, allowing the dynamic adaptation of the system to the prevailing environmental context.