Circular economy strategy based on digital technologies and the S4 framework: a case study on smartphones

Figures & data

Figure 1. Product life cycle description based on Deng et al. (2021).

Read the detailed description of this figure

The Figure describes the life cycle of electronic devices, from the acquisition of the raw material, the process, the assembly, the stage of satisfying the needs of each user, the opportunity to extend the life cycle of the devices, and the end of the device as it can be recycled, as well as becoming waste in its entirety.

Figure 2. Digital technology approach to reduce the path planning generator. (a) An unknown sample is approximate to a known sample. (b) A transfer learning approach reduces the number of operations. (c) A safe path is generated based on previous knowledge of virtual environment safe paths. Taken from (Maldonado-Romo, Maldonado-Romo, and Aldape-Pérez 2022).

Read the detailed description of this figure

The Figure describes an architecture for communicating a path generator between a known and unknown environment from the interaction between the domain change by a generative network. It describes samples from one house with known collisions and samples from a second house with unknown collisions to estimate the location of possible obstacles between different scenarios.

Figure 3. Path planning architecture. (a) Conventional path planning generator. (b) Reduced path planning architecture.

Figure 4. Generated path using virtual elements. (a) Path employs a paired approach. (b) Path employs unpaired approach. Taken from (Maldonado-Romo, Maldonado-Romo, and Aldape-Pérez 2022).

Read the detailed description of this figure

Figure describes a sample table showing the result of the route generated by the system. Two samples are tested with a scenario through samples associated one-to-one, the physical world with its virtual representation, and another scenario with samples that are not associated. A green path is generated to describe the collision-free path.

Table 1. Performance of this proposal and its standard deviation in 50 samples. Where $\downarrow$ Low is better, and $\uparrow$ Up is better. Taken from (Maldonado-Romo, Maldonado-Romo, and Aldape-Pérez 2022).

Model	Rel - std $\downarrow$	Rms - std $\downarrow$	$\mathit{lo}{\mathit{g}}_{\mathbf{10}}$ - std $\downarrow$	${\delta }_{\mathbf{1}}$ - std $\uparrow$	${\delta }_2$ - std $\uparrow$	${\delta }_{\mathbf{3}}$ - std $\uparrow$
Double-GAN-NR	0.189–0.011	0.433–0.009	0.108–0.015	0.792–0.020	0.897–0.016	0.976–0.021

Table 2. Path planning generator using augmented reality. Performance of the quantitative metrics in our approach and the standard deviation of 50 samples. $\downarrow$, lower is better; $\uparrow$, higher is better. Based on (Maldonado-Romo, Maldonado-Romo, and Aldape-Pérez 2022).

	Model	Accuracy Euclidean Distance (mean-std) $\downarrow$	Accuracy Manhattan Distance (mean-std) $\downarrow$	Accuracy Cosine Similarity (mean-std) $\downarrow$
Paired	Hierarchical clustering	12.6319 ± 1.6888	19.1410 ± 3.1404	0.9814 ± 0.01256
	Autoencoder	4.7478 ± 0.9862	7.4409 ± 1.2018	0.9814 ± 0.0224
Unpaired	Hierarchical clustering	2.1458 ± 0.7458	3.2458 ± 0.8756	0.8878 ± 0.0478
	Autoencoder	7.7589 ± 1.7895	11.4587 ± 2.4756	0.5478 ± 0.0789

Figure 5. Estimation of e-waste for smartphones.

Read the detailed description of this figure

The Figure depicts the behaviour of the devices considered as e-waste over some time in green, in blue the devices sold, and in red the devices that remained in stock.

Figure 6. DT a strategy toward lifetime extension and improved smartphone circular model.

Read the detailed description of this figure

The Figure describes the life cycle of electronic devices, from the acquisition of the raw material, the process, the assembly, the stage of satisfying the needs of each user, the opportunity to extend the life cycle of the devices, and the end of the device as it can be recycled, as well as becoming waste in its entirety. However, digital technologies are proposed to extend the functions of obsolete devices.

Table 3. Coefficient free collision. Based on (Maldonado-Romo, Maldonado-Romo, and Aldape-Pérez 2022).

	Model	Coefficient free collision
Paired	Hierarchical	0.92
	clustering
	Autoencoder	0.88
	Hierarchical	0.86
Unpaired	clustering
	Autoencoder	0.52

Deng, Q., M. Franke, E. S. Lejardi, R. M. Rial, and K.-D. Thoben 2021. Development of a Digital Thread Tool for Extending the Useful Life of Capital Items in Manufacturing Companies - An Example Applied for the Refurbishment Protocol. In 2021 26th ieee international conference on emerging technologies and factory automation (etfa), Västerås, Sweden, 1–8.

 Google Scholar

Maldonado-Romo, J., A. Maldonado-Romo, and M. Aldape-Pérez. 2022. “Path Generator with Unpaired Samples Employing Generative Adversarial Networks.” Sensors 22 (23): 9411. https://doi.org/10.3390/s22239411.

 PubMed Web of Science ®Google Scholar