EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Flexible Spectral Sensing Gripper for Real-Time Food Freshness Assessment

Yuhan Gong, Ruihua Zhang, Chunling Liu, Wei Liu, Wenjing Zhao, Yingle Du, Tao Sun, Xinqing Xiao

Eng · 16.05.2026 · 7 (5) · 243

DOI: 10.3390/eng7050243 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Eng, online erschienen am 16.05.2026. Verfasst von Yuhan Gong und 7 weiteren Autor:innen. Die Arbeit nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Reliable potato quality monitoring during postharvest handling requires compact sensing systems that can acquire chemically relevant information while operating on irregular tuber surfaces. In this study, a Flexible Spectral Sensing Gripper (FSSG) was developed by integrating a low-cost 12-channel visible/near-infrared (Vis/NIR) spectral sensor array, electronic components, and an ESP32-S microcontroller onto a flexible printed circuit (FPC) substrate encapsulated with PDMS. By embedding the sensing units into the grasping interface, the FSSG enables conformal, multi-point spectral acquisition during potato handling, reducing optical-coupling uncertainty associated with unstable contact. Spectral reflectance data were collected from potato tubers, and dry matter content (DMC) and starch content (SC) were determined by standard chemical analysis as reference values. Multiple linear regression (MLR) and partial least squares regression (PLSR) models were compared under Norm, SNV, MSC, SNV-Norm, and MSC-Norm preprocessing conditions, and support vector machine (SVM) classification was used to distinguish healthy and artificially induced deteriorated samples. Normalization combined with MLR provided the best performance among the evaluated regression approaches, achieving cross-validation coefficients of determination (RCV2) of 0.847 and 0.817 and RPD values of 2.557 and 2.345 for DMC and SC, respectively. The SVM model achieved 98.67% accuracy for healthy versus artificially induced deteriorated potato samples. Overall, the FSSG demonstrates the value of combining gripper-integrated spectral sensing with interpretable chemometric modeling for potato quality screening. The FSSG enables real-time non-destructive quality prediction and disease-detected classification of potatoes, improves sorting accuracy and production efficiency, and provides general sensing solutions for controlled-environment agriculture, cold-chain logistics, and value-added processing of agricultural products.

Zitieren

APA

Gong, Y., Zhang, R., Liu, C., Liu, W., Zhao, W., Du, Y., Sun, T., & Xiao, X. (2026). Flexible Spectral Sensing Gripper for Real-Time Food Freshness Assessment. Eng, 7(5), 243. https://doi.org/10.3390/eng7050243

Vancouver

Gong Y, Zhang R, Liu C, Liu W, Zhao W, Du Y, et al. Flexible Spectral Sensing Gripper for Real-Time Food Freshness Assessment. Eng. 2026;7(5):243. doi:10.3390/eng7050243

BibTeX
@article{Gong2026,
  author = {Gong, Yuhan and Zhang, Ruihua and Liu, Chunling and Liu, Wei and Zhao, Wenjing and Du, Yingle and Sun, Tao and Xiao, Xinqing},
  title = {Flexible Spectral Sensing Gripper for Real-Time Food Freshness Assessment},
  journal = {Eng},
  year = {2026},
  volume = {7},
  number = {5},
  pages = {243},
  doi = {10.3390/eng7050243}
}
Kostenlos starten →