Bio-inspired computation: Where we stand and what's next

Abstract

In recent years, the research community has witnessed an explosion of literature dealing with the mimicking of behavioral patterns and social phenomena observed in nature towards efficiently solving complex computational tasks. This trend has been especially dramatic in what relates to optimization problems, mainly due to the unprecedented complexity of problem instances, arising from a diverse spectrum of domains such as transportation, logistics, energy, climate, social networks, health and industry 4.0, among many others. Notwithstanding this upsurge of activity, research in this vibrant topic should be steered towards certain areas that, despite their eventual value and impact on the field of bio-inspired computation, still remain insufficiently explored to date. The main purpose of this paper is to outline the state of the art and to identify open challenges concerning the most relevant areas within bio-inspired optimization. An analysis and discussion are also carried out over the general trajectory followed in recent years by the community working in this field, thereby highlighting the need for reaching a consensus and joining forces towards achieving valuable insights into the understanding of this family of optimization techniques.

Introduction

Over millions of years, Nature has evolved to give rise to intelligent behavioral characteristics and biological phenomena, where adaptability, self-learning, robustness, and efficiency enable biological agents (such as insects and birds) to undertake complex tasks. While cases exemplifying these capabilities are truly multi-fold, the most illustrative ones revolve around the social behavior of animals such as ant colonies, beehives and bird flocks, where concepts such as stigmergy and the collective swarming movement of organisms often lead to the so-called Swarm Intelligence (SI), where improved exploration mechanisms over complex search spaces can be achieved by agents obeying local rules without any central control. The overall functionalities of the swarm are much richer than the simple sum of individual actions. Similarly, other renowned examples arise from the genetic inheritance process, the immune system of the human body or the neural activity of the brain. We refer to Ref. [1] for a comprehensive material summarizing these inspirational sources found in Nature.

Inspired by different behaviors observed in biological systems, many researchers in the research community investigating on computational paradigms have emulated intelligent bio-inspired processes in the form of computational algorithms, in an attempt to mimic the inherent advantages of such biological systems to address complex modeling, simulation, and optimization problems. In this regard, special attention has been paid to optimization problems, whose complexity has unleashed a rich substratum where to grow many bio-inspired population-based heuristic approaches, each differently balancing between computational efficiency and optimality of solutions. While the first contributions in this area are largely based on observation and emulation of Darwinian evolutionary principles, nowadays the number of bio-inspired solvers in the literature has increased dramatically, with very diverse inspirational rationale underneath their algorithmic design. This spotted flourishing of novel bio-inspiration optimization methods becomes even more intense when shifting the focus on other aspects related to optimization, such as multi-objective criteria, evolving (dynamic) optimization problems or distributed computing schemes, to mention a few.

However, quantity and diversity do not necessarily reflect scientific value when it comes to science. The development of the field has lately undergone a gold rush for bio-inspired streamlines that stimulate new algorithmic strands, around which some controversy has sprung regarding their relevance and novelty [2]. Debates around this topic are counterproductive, for which they waste efforts towards research directions with scarce – or even null – added scientific value. The same may occur in other research subareas as the ones exemplified above, where most algorithmic contributions build upon empirical performance observations rather than upon a deep, thoughtful and rigorous analysis of their design and internal operation. Futile debates should set aside to allow the entire community to start over with a clean common ground on the key research directions to be pursued in the future. We must ally to focus our efforts on important unresolved questions that can potentially produce greater insights into bio-inspired optimization techniques, ultimately leading to valuable advances and improved methods. Without a consensus, research niches of acknowledged relevance in the field will remain largely unexplored and unfairly dominated by controversial discussions, subtle and incremental algorithmic proposals, and a worrying lack of fresh breezes and fertile prospects.

This work responds to this need for a common meeting point by suggesting the audience to pause and reflect on which research directions should be pursued in the future in regards to bio-inspired optimization and related areas. For this purpose we pay special attention to numerical optimization, which was underneath the advent of the first bio-inspired solvers that were later adapted to combinatorial optimization. In this manuscript we provide an informed insight of the status of this field from both descriptive (where we stand) and prescriptive (what's next) points of view. This manuscript suggests and highlights several key research challenges that should captivate newcomers and experienced researchers for years to come, with scientific soundness at the core of their raison d'être. We hope that our envisioned future for bio-inspired computation acts as a suggestive guiding light for the community, bringing together different views that have remained so far quite different from each other to date, and potentially unifying them into a comprehensive multi-disciplinary view of the field.

The remainder of the paper is structured as follows: first, Section 2 provides a brief albeit informative overview of the history of bio-inspired computation. Section 3 and subsections therein undertake a comprehensive analysis of several areas of the field, stressing on their current status, trends, and open challenges. Section 4 elaborates on the general issues and research niches of bio-inspired computation, and finally Section 5 concludes this paper.