Machine Autonomy: Definition, Approaches, Challenges and Research Gaps

Ezenkwu, Chinedu Pascal; Starkey, Andrew

doi:10.1007/978-3-030-22871-2_24

Cited by 15 publications

(3 citation statements)

References 64 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In an attempt to provide a reasonable definition of machine autonomy, we previously categorised the attributes of autonomous agents into low-level and high-level attributes. This definition builds on a number of other definitions of the autonomous agent in the literature (Ezenkwu and Starkey 2019a).…”

Section: Machine Autonomymentioning

confidence: 99%

An unsupervised autonomous learning framework for goal-directed behaviours in dynamic contexts

Ezenkwu

Starkey

2022

Adv. in Comp. Int.

Self Cite

View full text Add to dashboard Cite

Due to their dependence on a task-specific reward function, reinforcement learning agents are ineffective at responding to a dynamic goal or environment. This paper seeks to overcome this limitation of traditional reinforcement learning through a task-agnostic, self-organising autonomous agent framework. The proposed algorithm is a hybrid of TMGWR for self-adaptive learning of sensorimotor maps and value iteration for goal-directed planning. TMGWR has been previously demonstrated to overcome the problems associated with competing sensorimotor techniques such SOM, GNG, and GWR; these problems include: difficulty in setting a suitable number of neurons for a task, inflexibility, the inability to cope with non-markovian environments, challenges with noise, and inappropriate representation of sensory observations and actions together. However, the binary sensorimotor-link implementation in the original TMGWR enables catastrophic forgetting when the agent experiences changes in the task and it is therefore not suitable for self-adaptive learning. A new sensorimotor-link update rule is presented in this paper to enable the adaptation of the sensorimotor map to new experiences. This paper has demonstrated that the TMGWR-based algorithm has better sample efficiency than model-free reinforcement learning and better self-adaptivity than both the model-free and the traditional model-based reinforcement learning algorithms. Moreover, the algorithm has been demonstrated to give the lowest overall computational cost when compared to traditional reinforcement learning algorithms.

show abstract

Section: Machine Autonomymentioning

confidence: 99%

An unsupervised autonomous learning framework for goal-directed behaviours in dynamic contexts

Ezenkwu

Starkey

2022

Adv. in Comp. Int.

Self Cite

View full text Add to dashboard Cite

show abstract

“…There is no definitive definition of what an autonomous machine is [3], [4]. Autonomy means a self-governing state [5]; thus, an autonomous machine implies a self-governing apparatus.…”

Section: Introductionmentioning

confidence: 99%

Developmental Autonomous Behavior: An Ethological Perspective to Understanding Machines

Isaka

2023

IEEE Access

View full text Add to dashboard Cite

Developmental autonomous behavior refers to the general ability of a machine to acquire new skills and behavior from its birth to maturity on its own without human intervention. This article describes the principles of behavior development in machines, providing a practical framework to analyze and synthesize machines with developmental capabilities. Inspired by biological views of behavioral causation, the work emphasizes principled explanations to, not only the "how" question on mechanisms but also the "why" question on causation of behavior development. This ethology-oriented perspective offers a renewed opportunity to construct a theoretical framework from the ground up, overcoming the age-old problems of intrinsic motivation and symbol emergence in autonomous machines. One of the key contributions of this article is the logical explanation of why and how value systems drive successive development of memory functions, resulting in progressive changes in behavior from innate reflexive to episodic, procedural, and autonomic behavior. Another notable contribution is the logical and plausible explanation of why and how a physical sensorimotor system evolves to manipulate its internal memory, fostering conceptual and social behavior development. This article provides an extensive review of prior research, followed by detailed descriptions of the causality and mechanisms of behavior development, and concludes with discussions on criticism, future work, ethics, and system architecture.

show abstract

“…Leveraging on this potential, unsupervised techniques have been used to confine continuous space observations and/or actions to a finite space [5], [6], [7], [8], hence simplifying an agent's world for real-time exploitation. Yet, traditional unsupervised learning methods are unable to provide an agent with the right behaviour information since the sensory information is not mapped to specific outputs or actions as the case may be [9], [10]. Due to this limitation, these methods are often combined, to act as vector quantisation techniques, with supervised learning [7] or reinforcement learning algorithms [11].…”

Section: Introductionmentioning

confidence: 99%

Unsupervised Temporospatial Neural Architecture for Sensorimotor Map Learning

Ezenkwu

Starkey

2021

IEEE Trans. Cogn. Dev. Syst.

Self Cite

View full text Add to dashboard Cite

The ability to learn the sensorimotor maps of unknown environments without supervision is a vital capability of any autonomous agent, be it biological or artificial. An accurate sensorimotor map should be able to encode the agent's world and equip it with the capability to anticipate or predict the results of its actions. However, to design a robust autonomous learning technique for an unknown, dynamic, partially observable or noisy environment remains a daunting task. This paper proposes a Temporospatial Merge Grow When Required (TMGWR) network for continuous self-organisation of an agent's sensorimotor awareness in noisy environments. TMGWR is an adaptive neural algorithm that learns the sensorimotor map of an agent's world using a time series self-organising strategy and the Grow When Required (GWR) algorithm. The algorithm is compared with GNG, GWR and TGNG in terms of their disambiguation performance, sensorial representation accuracy and sensorimotor-link error, a new metric that is developed in this paper to evaluate how well a sensorimotor map represents causality in the agent's world. The outcomes of the experiments show that TMGWR is more efficient and suitable for sensorimotor map learning in noisy environments than the competing algorithms.

show abstract

Machine Autonomy: Definition, Approaches, Challenges and Research Gaps

Cited by 15 publications

References 64 publications

An unsupervised autonomous learning framework for goal-directed behaviours in dynamic contexts

An unsupervised autonomous learning framework for goal-directed behaviours in dynamic contexts

Developmental Autonomous Behavior: An Ethological Perspective to Understanding Machines

Unsupervised Temporospatial Neural Architecture for Sensorimotor Map Learning

Contact Info

Product

Resources

About