Harnessing Patterns to Support the Development of Hybrid Quantum Applications

TL;DR

This paper introduces an automated approach for detecting, selecting, and aggregating patterns to streamline the development of hybrid quantum applications, reducing manual effort and expertise needed.

Contribution

It presents a novel method that automates pattern detection, implementation selection, and solution aggregation for hybrid quantum application development.

Findings

Automated pattern detection for quantum applications

Automated selection of implementations based on requirements

Aggregation of solutions into executable quantum programs

Abstract

Quantum computing provides computational advantages in various domains. To benefit from these advantages complex hybrid quantum applications must be built, which comprise both quantum and classical programs. Engineering these applications requires immense expertise in physics, mathematics, and software engineering. To facilitate the development of quantum applications, a corresponding quantum computing pattern language providing proven solutions to recurring problems has been presented. However, identifying suitable patterns for tackling a specific application scenario and subsequently combining them in an application is a time-consuming manual task. To overcome this issue, we present an approach that enables (i) the automated detection of patterns solving a given problem, (ii) the selection of suitable implementations fulfilling non-functional requirements of the user, and (iii) the automated aggregation of these solutions into an executable quantum application.