Synchronized Object Detection for Autonomous Sorting, Mapping, and Quantification of Materials in Circular Healthcare

TL;DR

This work tackles the challenge of real-time material monitoring for a circular economy by introducing a synchronized, multi-unit object-detection framework that enables autonomous sorting, mapping, and quantification. It formalizes a thermodynamical material network (TMN) model and couples it with material measurement units (MMUs) that provide synchronized stock estimates across distributed vision units, yielding synchromaterials. The authors demonstrate the approach with a numerical TMN example and a medical-material prototype using inhalers, including a dedicated 5IPP inhaler dataset and a distributed detection setup that processes at real-time frame rates. The results show the feasibility of real-time, wide-area material monitoring and sorting, with public dataset, code, and demos, highlighting the potential to enhance decision-making in circular-material systems.

Abstract

The circular economy paradigm is gaining interest as a solution to reducing both material supply uncertainties and waste generation. One of the main challenges in realizing this paradigm is monitoring materials, since in general, something that is not measured cannot be effectively managed. In this paper, we propose a real-time synchronized object detection framework that enables, at the same time, autonomous sorting, mapping, and quantification of solid materials. We begin by introducing the general framework for real-time wide-area material monitoring, and then, we illustrate it using a numerical example. Finally, we develop a first prototype whose working principle is underpinned by the proposed framework. The prototype detects 4 materials from 5 different models of inhalers and, through a synchronization mechanism, it combines the detection outputs of 2 vision units running at 12-22 frames per second (Fig. 1). This led us to introduce the notion of synchromaterial and to conceive a robotic waste sorter as a node compartment of a material network. Dataset, code, and demo videos are publicly available.

Figures (7)

• Figure 1: High-level summary of the paper: real-time inhaler detection enables both material sorting, mapping, and quantification when it is synchronized with other detection units.
• Figure 2: Analogy between a material network (left) and an electrical network (right). Wide-area monitoring of these networks is performed via material measurement units (MMUs) zocco2022material and PMUs khajeh2015integrated, respectively.
• Figure 3: Material network considered in the example and the prototype.
• Figure 4: Synchromaterials of the numerical example.
• Figure 5: Training samples ((a)--(d)) and tracking results ((e)--(h)).

Theorems & Definitions (3)

• Definition 1: bondy1976graph
• Definition 2: zocco2023thermodynamical
• Definition 3: Synchromaterial