ZLP-Perception: Zero-Local-Provisioning of Perceptual Capabilities in Heterogeneous Systems
Abstract
Perception is commonly implemented through sensors, sensing modules, and locally deployed perception algorithms attached to individual devices, machines, vehicles, robots, or computational systems. Although local perception can provide low latency, operational independence, and direct access to environmental information, requiring every object to possess its own complete set of perceptual capabilities may result in duplicated sensing infrastructure, redundant computation, unnecessary maintenance, and inefficient utilization of shared observation resources. In heterogeneous systems, some perceptual requirements may instead be satisfied through measurements, representations, or perception services provided by other objects or shared infrastructure.
This paper proposes ZLP-Perception, a domain-specific conceptual framework derived from Zero-Local-Provisioning (ZLP) for investigating the collective provision of perceptual capabilities. Its central principle is that the local provisioning quantity of a selected perceptual trait may be exactly zero for an individual object, provided that the system continues to satisfy the associated functional requirements through compatible external observations, shared sensing resources, remote perception services, alternative sensing modalities, or other permitted mechanisms. The absence of a local sensor or perception capability is therefore treated as an explicit configuration decision rather than an automatic loss of system functionality.
The framework distinguishes local sensing hardware, local perception computation, externally supplied observations, effective perceptual availability, and downstream task requirements. It formalizes objects, perceptual traits, observation requests, sensing providers, spatiotemporal coverage, information quality, uncertainty, service latency, reliability, privacy, and lifecycle costs. A joint optimization formulation determines which perceptual capabilities should be provisioned locally and which requirements can be satisfied collectively. The formulation accommodates binary sensor-presence decisions, continuous sensing capacity, alternative modalities, and task-dependent service constraints.
ZLP-Perception does not assume that perception can always be externalized. Local sensing may remain indispensable when observations are inaccessible to other providers, communication latency is unacceptable, or safety and reliability requirements demand independent operation. This paper develops the conceptual model, derives feasibility conditions, presents a reference architecture and algorithmic template, and proposes an empirical evaluation methodology. The framework is presented as a research proposal whose novelty, generality, and practical benefits require systematic comparison with established approaches and experimental validation.
Related
- ZAC-Perception: Collective Provisioning of Perceptual Capabilities in Multi-Agent Embodied Intelligence
- ZLP-Inference: Zero-Local-Provisioning of Inference Capabilities in Heterogeneous Intelligent Systems
- Zero-Local-Provisioning for Training (ZLP-Training): A General Framework for Collective Provisioning of Training Capabilities
- Beyond Individual Large Language Models: A General Framework for Intelligent and Autonomous Provisioning of AI Capabilities
- Zero-Local-Provisioning (ZLP): A General Framework for Collective Provisioning of Resources, Capabilities, and Attributes
BibTeX
@misc{zhang2026zlpperception,
title = {{ZLP-Perception: Zero-Local-Provisioning of Perceptual Capabilities in Heterogeneous Systems}},
author = {Jincheng Zhang},
howpublished = {Zenodo},
year = {2026},
month = oct,
doi = {10.5281/zenodo.23451096},
url = {https://zenodo.org/records/23451096},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - ZLP-Perception: Zero-Local-Provisioning of Perceptual Capabilities in Heterogeneous Systems
AU - Jincheng Zhang
PY - 2026
DA - 2026/10/18
PB - Zenodo
DO - 10.5281/zenodo.23451096
UR - https://zenodo.org/records/23451096
AB - Perception is commonly implemented through sensors, sensing modules, and locally deployed perception algorithms attached to individual devices, machines, vehicles, robots, or computational systems. Although local perception can provide low latency, operational independence, and direct access to environmental information, requiring every object to possess its own complete set of perceptual capabilities may result in duplicated sensing infrastructure, redundant computation, unnecessary maintenance, and inefficient utilization of shared observation resources. In heterogeneous systems, some perceptual requirements may instead be satisfied through measurements, representations, or perception services provided by other objects or shared infrastructure. This paper proposes ZLP-Perception, a domain-specific conceptual framework derived from Zero-Local-Provisioning (ZLP) for investigating the collective provision of perceptual capabilities. Its central principle is that the local provisioning quantity of a selected perceptual trait may be exactly zero for an individual object, provided that the system continues to satisfy the associated functional requirements through compatible external observations, shared sensing resources, remote perception services, alternative sensing modalities, or other permitted mechanisms. The absence of a local sensor or perception capability is therefore treated as an explicit configuration decision rather than an automatic loss of system functionality. The framework distinguishes local sensing hardware, local perception computation, externally supplied observations, effective perceptual availability, and downstream task requirements. It formalizes objects, perceptual traits, observation requests, sensing providers, spatiotemporal coverage, information quality, uncertainty, service latency, reliability, privacy, and lifecycle costs. A joint optimization formulation determines which perceptual capabilities should be provisioned locally and which requirements can be satisfied collectively. The formulation accommodates binary sensor-presence decisions, continuous sensing capacity, alternative modalities, and task-dependent service constraints. ZLP-Perception does not assume that perception can always be externalized. Local sensing may remain indispensable when observations are inaccessible to other providers, communication latency is unacceptable, or safety and reliability requirements demand independent operation. This paper develops the conceptual model, derives feasibility conditions, presents a reference architecture and algorithmic template, and proposes an empirical evaluation methodology. The framework is presented as a research proposal whose novelty, generality, and practical benefits require systematic comparison with established approaches and experimental validation.
ER -