Conference Agenda

This work combines laboratory experiments, dynamic modeling, and a preliminary scale-up analysis to evaluate the feasibility

of using coconut water as a substrate for producing Saccharomyces cerevisiae biomass. The process was modeled as a batch reactor using a dynamic system in , based on biomass and substrate balances under Monod kinetics.

Experimental data were collected during 10 h of cultivation (n = 6 time points). The observed specific growth rate in the exponential phase was estimated using linear regression of , obtaining with .

This value is significantly lower than those reported for optimized synthetic media, suggesting kinetic limitations due to the natural, non- supplemented substrate.

The final biomass concentration reached was , which limits volumetric productivity. Since residual sugar was not measured, the yield was taken from literature to complete the model.

A preliminary design for a pilot plant targeting requires about of operating volume (approximately total with headspace). The low biomass concentration shifts the bottleneck to downstream processing, highlighting the importance of estimating real kinetic parameters before industrial scale-up.

The discharge of synthetic dyes into aquatic systems represents a significant environmental challenge due to their persistence and resistance to degradation. This study investigates the sustainable valorization of avocado seed biomass through hydrothermal carbonization (HTC) for methylene blue removal from water. A Box–Behnken experimental design was applied to evaluate the effects of temperature (180–250 °C), residence time (2–8 h), and solid-to-water ratio (1:5–1:15) on adsorption performance. The results indicated that temperature was the most influential factor, with optimal conditions identified at 180 °C, 5 h, and a solid-to-water ratio of 1:15. To further enhance adsorption capacity, deep eutectic solvents (DES, choline chloride–urea) and natural deep eutectic solvents (NADES, choline chloride–thiourea) were incorporated during HTC at different molar ratios. DES treatment significantly improved dye removal, achieving the highest efficiency at a 1:2 molar ratio, while NADES showed moderate enhancement. SEM, BET, and FTIR analyses confirmed that eutectic-assisted HTC influenced surface morphology, mesoporosity, and functional group distribution. The findings demonstrate that DES-assisted hydrothermal carbonization of avocado seed represents an environmentally friendly strategy for improving hydrochar adsorption performance, contributing to biomass valorization and sustainable water treatment technologies.

Current School Emergency Plans (PISE) in Chile rely on standardized risk assessments and static maps to design evacuation protocols. This approach fails to account for real-world logistical constraints critical for vulnerable coastal populations, such as heterogeneous crowds, street capacity, varying topography, and latency times of institutional alert systems. This Work in Progress research proposes a quantitative framework using graph theory and logistics optimization to transition PISE from administrative compliance to dynamic operational management. Evacuation routes are modeled as networks, introducing velocity penalties based on terrain slope and student mobility, constrained by a strict survival window equation. Preliminary results from Cartagena, cross-referenced with official 2025 regional drill data, validate the model: standard drill times are identified as "false positives" achieved under controlled traffic conditions. To resolve this, the project introduces DERA (Delayed Evacuation Routing Algorithm), a software architecture proposing staggered departures to eliminate multicommodity flow bottlenecks, theoretically increasing the arrival rate of students to safety zones within the survival window and significantly reducing the probability of system collapse during real-world evacuation scenarios.

Residues from fruits are a sustainable source of valuable products for the food industry. For instance, pectin is a polysaccharide that is widely used as an additive and it also could be used as a film former polymer for food packaging. In Ecuador, residues of babaco (Carica pentagona) are generated, but they are not currently used in the industry. In this work, two methods of pectin extraction from babaco residues were compared and active capacity of films obtained from extracted pectin was evaluated. The efficiency of the methods was compared by a multifactorial analysis of the extraction field. The active capacity of pectin films was evaluated through antioxidant compound release assay, using a food simulant. The ultrasound-assisted extraction method (UAE) showed a higher extraction field of pectin than the acid hydrolysis. Moreover, the amplitude of UAE and the type of residue influenced significantly in the pectin extraction field. The antioxidant compound release from the pectin film to the food simulant reached the equilibrium after three hours. The kinetics were adjusted to Korsmeyer–Peppas model where the kinetic parameters showed a Fickian behavior without affecting the material structure. These results show the potential of babaco residues as an alternative source of pectin and its application for the development of active food packaging materials.

NEREO is a Remotely Operated Vehicle (ROV) developed as an academic prototype for an Embedded Systems course. This compact, low-cost ROV is designed for the automated monitoring of environmental parameters in fish farms, measuring pH, water temperature using a DS18B20 sensor, and orientation variables through an MPU6050 IMU. Its structure was designed in SolidWorks and manufactured in PLA using 3D printing. It features a four-motor propulsion system in an “X” configuration, providing precise maneuverability to operate at depths of up to 2 meters.

In terms of embedded hardware, it integrates a reliable sensor architecture centralized in a processing unit that remotely communicates data via the MQTT protocol to the ThingsBoard Demo platform, where monitored variables can be visualized in real time and motors can be controlled from a customized dashboard. Additionally, it incorporates proportional control with IMU feedback to dynamically stabilize the vehicle, improving its performance against inclinations and disturbances during supervised aquatic testing.

The project concludes with the complete integration of design, electronics, software, and communications, demonstrating robust functional performance. NEREO positions itself as an effective platform for environmental monitoring in fish farms and as a foundation for future improvements in autonomous navigation, encapsulation, and expansion of the sensory system.

The widespread detection of fluoroquinolones in aquatic environments represents a growing environmental challenge due to their persistence and potential contribution to antimicrobial resistance. This work reports the development of a sustainable magnetic biochar@ZIF-8 composite derived from almond shell biomass for the efficient removal of fluoroquinolones from water. Biochar production was optimized using a central composite design evaluating pyrolysis temperature and residence time to maximize yield and adsorption performance. The optimized biochar was chemically activated with KOH, magnetized through in situ Fe₃O₄ formation, and subsequently modified via in situ growth of ZIF-8, generating a hierarchically porous structure with enhanced surface area and functional functionalities. Characterization by SEM–EDS, FTIR, BET, and TGA confirmed the successful integration of magnetic and metal–organic phases. Adsorption experiments demonstrated that removal efficiency was highest at pH 6, where fluoroquinolones predominantly exist in zwitterionic form. Kinetic studies were best described by the pseudo-second-order model, suggesting chemisorption mechanisms governed by π–π interactions, hydrogen bonding, and coordination with Zn²⁺ active sites. Furthermore, the composite retained more than 84% of its adsorption capacity after four regeneration cycles and enabled rapid magnetic separation. The results highlight the potential of almond shell-derived magnetic biochar@ZIF-8 as a cost-effective and environmentally sustainable material for water remediation applications.

This paper presents Twin5G, a proposed intelligent digital twin for heterogeneous 5G networks (HetNet) designed for advanced monitoring, prediction, operational automation, and evolution toward production environments.

The system integrates initial network simulation, cloud-based analytical storage, and artificial intelligence to represent, analyze, and anticipate infrastructure behavior before critical events occur in the real network.

The proposed architecture considers a heterogeneous network composed of macrocells, microcells, picocells, and femtocells, along with an intelligent processing layer capable of supporting tasks such as congestion prediction, latency estimation, anomaly detection, decision automation, and resource self-management. In its current state, Twin5G implements
an initial simulation phase in MATLAB Online to generate
performance metrics such as latency, throughput, received power, and link conditions, which are then exported for
processing in the cloud.
As an architectural evolution, the use of Google
Colab for data preparation, BigQuery as the analytical foundation
of the digital twin, Vertex AI as an intelligent layer for inference
and predictive analytics, and Cloud Run as the deployment environment for
an interactive web interface is proposed. Additionally, Looker
Studio can be used for executive visualization of historical
metrics and results.

Twin5G is conceived as a scalable foundation for building intelligent management systems in 5G networks, capable of supporting advanced monitoring functions, operational optimization, and proactive decision-making in heterogeneous, high-demand scenarios.

Securing wireless networks has become a critical challenge in cybersecurity, as attackers can exploit vulnerabilities to steal credentials and private information. This project developed a hybrid intrusion detection framework designed to enhance wireless network monitoring through behavioral analysis. The system integrates Zeek, a network telemetry framework, with Suricata, a signature-based intrusion detection system, on a Raspberry Pi configured as a wireless gateway. Logs generated by these tools were used to extract behavioral features and train Machine Learning anomaly detection model. Ultimately, results indicate that integrating traditional IDS mechanisms with AI-driven anomaly analysis strengthens adaptive wireless network security.