Executive Summary

February 10, 2026

Emerging Technologies in Mental Health Monitoring of App Drivers

Sabrina Johansen Tadei; Ingrid de Matos Martins

Summary prepared by the ResumeAI tool, an artificial intelligence solution developed by the Pecege Institute focused on synthesis and writing.

The objective of this research was to investigate how emerging technologies can contribute to the monitoring and psychological support of app drivers in Brazil, considering working conditions with long hours, low pay, and emotional overload that impact their mental health. The study sought to identify the main stress factors, assess drivers’ receptiveness to digital solutions for well-being, and explore the opportunities and challenges in implementing tools such as mindfulness apps and Artificial Intelligence-based systems. The investigation started from the premise that, although digital platforms have revolutionized urban mobility, the well-being of the workers who sustain this ecosystem remains in the background, creating a gap that technology, applied ethically, could help fill.

The labor market has been reconfigured by the platform economy, a phenomenon known as “uberization”. This app-mediated model alters traditional employment relationships (Anwar & Graham, 2020). Embedded in the fourth industrial revolution, which integrates technologies such as Big Data and IoT, this paradigm dissolves the boundaries between the physical and digital domains (Antunes, 2020). The term “uberization”, originating with the rise of Uber, describes a labor organization that, under the banner of flexibility, often masks precarious conditions and algorithmic control (Abílio et al., 2021).

Studies associate gig work with high levels of psychological distress. Research with app drivers reveals emotional exhaustion, weakening of social bonds, and difficulty in recognition, factors that negatively impact mental health (Ferreira et al., 2023; Sousa, 2024). The pressure for productivity, financial instability, and the absence of institutional support create an environment conducive to disorders such as anxiety, depression, and chronic stress. This reality is aggravated by the solitary nature of the work and exposure to risks in traffic and in interactions with passengers.

In parallel, emerging technologies arise as a strategy to mitigate mental health problems. Solutions such as mindfulness applications, vital signs monitoring devices, and AI systems for stress detection have been proposed to offer remote psychological support and reduce occupational stress (Campos et al., 2024). However, implementation faces barriers such as data privacy, digital access inequality, user resistance, and lack of regulation, which require ethical and institutional attention for their effective use (Filipak et al., 2024; Freitag et al., 2023).

In Brazil, the relevance of the topic is amplified by socioeconomic data. In 2022, the IBGE indicated that 1.5 million Brazilians worked through apps, with 778,000 dedicated to passenger transport. This contingent faces a struggle for basic rights, such as benefits, safety, and regulated rest, in a scenario of uncertainty that affects their well-being (Weege et al., 2024). The high unemployment rate drives adherence to this work modality, aggravating the category’s vulnerability (Candido, 2020). It becomes imperative to investigate the potential of emerging technologies as a support tool, filling a gap in the literature that still incipiently explores the application of these solutions with these workers (Melo et al., 2022).

The research had an applied nature, a descriptive objective, and a mixed approach, combining qualitative and quantitative methods for a comprehensive analysis. The target population consisted of active app drivers in Brazil, over 18 years old, operating on platforms such as Uber, 99, and Indrive, with a focus on major urban centers. Participation was voluntary and conditioned on free and informed consent, in accordance with the General Data Protection Law (Law No. 13.709/2018), ensuring the confidentiality of information.

Data collection was carried out through an online survey, with a hybrid questionnaire composed of structured and unstructured questions, based on previous studies on platform work, mental health, and technology (Silva, 2024; Alves et al., 2021). The structured questions included measurement scales for quantitative analysis, inspired by validated instruments (Almeida et al., 2022), allowing the collection of data on stress frequency and work schedule. The open-ended questions were designed to capture individual experiences, exploring working conditions, emotional state, and receptivity to new technologies.

The analysis of quantitative data was conducted using descriptive statistics with Jamovi software to calculate frequencies, percentages, means, and standard deviations, focusing on pattern identification. For qualitative data, Bardin’s content analysis was employed, following the stages of pre-analysis, material exploration, and results treatment. Thematic categorization resulted in central categories such as occupational stress, perception of exploitation, impacts on family life, absence of emotional support, and receptivity to the use of technologies for mental health care.

The study has limitations. The sample of 101 respondents was non-probabilistic, by convenience, with a majority concentration of drivers from São Paulo and of the male gender. Consequently, the results are not generalizable and should be interpreted as exploratory indicators. The cross-sectional design, based on self-report, captures a specific moment and is subject to biases, not allowing the establishment of causal inferences. The absence of inferential statistical analyses also limits the depth of the conclusions.

The profile of the 101 participants is predominantly cisgender men (89.1%), with an average age of 36.5 years, a productive age range associated with pressures for financial stability (Dejours, 2020). The majority (63.1%) work in the city of São Paulo. Regarding years of experience, 36.0% have been working in the activity for between 1 and 3 years, suggesting a concentration of professionals with intermediate experience.

A central finding relates to working hours: 47.5% of drivers reported working between 9 and 12 hours per day, while 14.9% exceed 12 hours daily. Combined, more than 62% of the sample are subjected to extended working hours. This reality reinforces the thesis that the flexibility of app-based work translates into a self-imposition of long hours to achieve a viable income, transferring the responsibility for their livelihood entirely to the worker (Silva and Henning, 2022).

The correlation between working hours and the impact on mental health is evident: 79.3% of drivers feel stressed with some regularity, with 34.7% reporting stress “frequently” and 44.6% “sometimes”. When asked if their work routine affects their mental health, 68.3% responded affirmatively. Among the negative effects, stress (32.3%) leads, followed by the perception of exploitation (27.7%) and anxiety (13.8%). The weariness transcends work, with reports of interference in personal life. One respondent illustrated: “I arrive so tired that I can’t even talk to my family properly. I just want to switch off.”

Despite the wear and tear, most drivers do not intend to abandon the activity, a paradox anchored in financial necessity and the difficulty of reinsertion into the formal market. Remaining in the activity highlights a complex relationship of dependence, even with the recognition of psychological harm (Oliveira and Coutinho, 2021). This scenario is reinforced by the perception that, although self-employed, they are subject to the control of platform algorithms, which dictate prices and routes, configuring new forms of subordination (Brasil, 2021).

The data reveal a gap between the perception of the problem and the search for help. Only 18.8% of drivers who feel affected sought psychological help in the last 12 months. This low rate can be attributed to stigma, costs, and lack of time (Kohn et al., 2021). Most drivers are unaware of the mental health programs offered by the platforms, and the charging for these services exacerbates the access barrier, contradicting recommendations from the World Health Organization for precarious populations (World Health Organization, 2020).

In contrast, drivers showed remarkable receptiveness to the use of technologies for mental health care. The majority (61.5%) stated they had already used some digital resource for this purpose. When asked about what types of technology could assist, applications were the most mentioned option (44.4%), followed by online therapy (22.2%). This data suggests a fertile ground for the implementation of tools such as chatbots and fatigue sensors. According to Honorato et al. (2025), AI-based technologies can offer resources for remote emotional support. The analysis also revealed that drivers with less than three years of experience showed a greater propensity to use these tools.

However, the adoption of monitoring technologies raises concerns for 27.5% of participants. The main concern was data privacy (44.4%), followed by lack of time to use the technologies (27.8%) and complexity of use (22.2%). One driver summarized the fear: “I’m afraid my data will be used against me.” This perception highlights the need for any initiative to be implemented with transparency and security, preserving ethical principles (Mamykina et al., 2022). The balance between monitoring for well-being and corporate surveillance is crucial for the tools to be perceived as support, rather than control (Torous et al., 2020).

Drivers’ perception of the platforms is mostly critical. The data show that 72.3% of respondents agree that there is a lack of emotional support from the companies. A testimonial illustrates this frustration: “The companies don’t care about our mental exhaustion, only about the number of rides.” This view is corroborated by the belief of 51.5% of drivers that mental health monitoring should be an obligation of the companies themselves. Although the platforms have consolidated through innovation, this potential has not been directed towards the care of human capital, generating the perception that the focus lies on profit (Ravanelle, 2019).

The results confirmed that the work routine of app drivers significantly compromises mental health, resulting in high levels of stress and anxiety. However, the study revealed a considerable openness to the use of digital solutions, such as mindfulness apps and AI-monitored break systems. These tools are perceived as viable resources to support psychological care, provided their implementation is guided by ethics and data security. The research has limitations, such as the concentration of the sample in São Paulo, which restricts the generalization of findings, and its cross-sectional design based on self-report, which prevents causal inferences. The analysis, limited to descriptive statistics, positions the results as exploratory.

For future investigations, it is recommended to expand the sample to cover different regions of Brazil, adopt longitudinal studies, and conduct intervention research to test the effectiveness of specific emerging technologies. It is relevant to explore the potential of Big Data and Artificial Intelligence to identify early signs of mental illness, as well as to deepen the investigation into the ethical, privacy, and regulatory aspects of digital tools. It is concluded that the objective was achieved: it was demonstrated that emerging technologies, when applied with ethics and institutional support, present significant potential to promote mental health and improve the quality of life of app drivers in Brazil.

References:
ABÍLIO, Ludmila C.; BELLINI, Rafaela; SANTOS, Mariana M. Asymmetric partners: work and health of app drivers. Ciência & Saúde Coletiva, Rio de Janeiro, v. 26, n. 12, p. 5915-5924, Dec. 2021.
ALMEIDA, A. N. de; PEREIRA, M. R.; SANTOS, F. L. Using the Technology Acceptance Model to Characterize Barriers and Opportunities of Telemedicine in Rural Populations: Survey and Interview Study. Journal of Medical Internet Research, v. 24, n. 4, e35436207, 2022.
ALVES, Aline et al. Mental health and precariousness: well-being of app drivers. Universidade Federal do Rio Grande do Sul, 2021.
ANTUNES, Ricardo. Uberization, digital work, and industry 4.0. São Paulo: Boitempo, 2020.
BRASIL, P. M. A. Digital platforms and app-based work: new forms of control and organization. Revista do Tribunal Superior do Trabalho, Brasília, v. 87, n. 1, p. 87–109, Jan./Mar. 2021.
CAMPOS, Israel Marques; SOARES, Anna Carolina Guedes; FROEDE, Clara de Souza; et al. The role of digital technologies in promoting the mental health of healthcare professionals. Humanidades e Tecnologia (FINOM), v. 31, n. 1, 2024.
CANDIDO, L. A. C. Psychic, social, and political impacts of uberization. 2020. Undergraduate Thesis (Graduation in Psychology) – Universidade Federal de Mato Grosso do Sul, Campo Grande, 2020.
DEJOURS, Christophe. Psychodynamics of work: suffering and mental health. São Paulo: Cortez, 2020.
FERREIRA, Fabrício Gonçalves; MENDES, Elzilaine Domingues; NAVES, Emilse Terezinha. Malaise in uberization: reflections on work from the perspective of neoliberal logic. Cadernos de Psicologia Social do Trabalho, São Paulo, v. 26, n. 2, 2023.
FILIPAK, Larissa Elisa; ARAÚJO, Marília Daniella Machado; et al. Digital technology in mental health: limits and potential. Revista da Unidade de Investigação do Instituto Politécnico de Santarém, v. 12, n. 1, 2024.
FREITAG, Abel Felipe; ARDENGUE, Mariana; MARCHINI, Kauana Borges; et al. Technological tools used to improve the physical and mental health of the population: a systematic review. Arquivos de Ciências da Saúde da UNIPAR, v. 27, n. 1, 2023.
HONORATO, E. J. S. et al. Mental health in the post-Covid-19 era: exploring the contribution of artificial intelligence in a literature review. Revista Foco, São Paulo, v. 18, n. 7, p. 1-25, 2025.
INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICA (IBGE). Unemployment. Brasília: IBGE, 2025.
KOHN, R.; ALI, A. A.; PUAC-POLANCO, V.; FIGUEROA, C.; LÓPEZ-SOTO, V.; MORGAN, K. et al. Barriers to mental health care: global perspectives. The Lancet Psychiatry, London, v. 8, n. 11, p. 983–992, 2021.
MAMYKINA, L.; SMALDONE, A.; BAKKEN, S. Ethical considerations in the use of big data for mental health research. Journal of Medical Internet Research, Toronto, v. 24, n. 3, e34567, 2022.
MELO, a. S. O. et al. Navigating virtual waves: barriers and facilitators to digital inclusion. Research, Society and Development, Lisbon, v. 11, n. 9, e19111931685, 2022.
OLIVEIRA, E. L. de; COUTINHO, M. C. Psychic suffering and mental health at work: new configurations under digital capitalism. Revista Psicologia e Sociedade, Belo Horizonte, v. 33, n. 1, p. e020017, 2021.
RAVANELLE, Alexandrea J. Hustle and gig: struggling and surviving in the gig economy. Oakland: University of California Press, 2019.
SILVA, A. P.; HENNING, C. Flexibilization and intensification of work on digital platforms: impacts on psychic health. Revista Brasileira de Saúde Ocupacional, São Paulo, v. 47, p. 1-13, 2022.
SOUSA, Ana Laís Carvalho de. Experiences of uberized work and mental health of food delivery app drivers. 2024. Master’s Thesis (Psychology) – Universidade Federal do Ceará, Fortaleza, 2024.
TOROUS, John; NICHOLAS, Jennifer; LARSEN, Mark E.; FIRTH, Joseph; CHRISTENSEN, Helen. Digital mental health and COVID-19: using technology today to accelerate the curve on access and quality tomorrow. JMIR Mental Health, Toronto,


Executive summary from the Final Course Work of Specialization in Digital Business from the MBA USP/Esalq

Learn more about the course; click here:

Who edited this article

Most recent

You may also like

November 12, 2026

Business plan for a company specializing in custom skins for DJ equipment

The market for DJ equipment has shown strong growth in recent years, driven by the advancement of digital entertainment and music culture, which revealed a gap in Brazil for the protection and customization of these high-cost items. Protective films emerged as a solution to simultaneously meet the need for conservation and customization. The objective of the study was to develop a business plan for a company specializing in customized films for DJ equipment, evaluating its market and financial viability. The methodology adopted was based on analyses of publicly accessible secondary data, including market trends, international benchmarking, competitor analysis, and the construction of financial projections. The results indicated a partially unmet existing demand and opportunities for differentiation through greater customization options and better strategic positioning via social media. The financial analysis demonstrated the economic viability of the project, with profit generation in the first year of operation, secure investment, and a comfortable profit margin on the product. It was concluded that the creation of the company is viable and financially advantageous, with low initial investment costs, provided it focuses on differentiation, product variety, and adequate market strategies.

Keywords: Strategic analysis; Customization; Entrepreneurship; Online retail; Economic viability.

November 12, 2026

Financial fraud in Brazil under an analytical perspective with public data

The scenario of expansion of instant payment methods and the consolidation of fintechs in Brazil have intensified users’ exposure to digital financial fraud. The study aimed to map the use of the main financial transactions, evaluate the scenario of fraud, disputes, and chargebacks, and relate these findings to corporate governance and compliance practices. The methodology was descriptive, with a mixed approach, and used official secondary data from the Central Bank, CVM, and Febraban, in the period from 2023 to 2025. Business Intelligence and data visualization techniques were applied with Power BI to analyze the volume of transactions, especially via Pix, and identify patterns of disputes and chargebacks through the Special Chargeback Mechanism (MED). The results revealed the expressive growth of Pix in quantity and value, the accelerated increase in disputes, and a notable discrepancy between the amounts accepted and those effectively returned via MED. An increase in the number of users with fraud markings was also observed. It was concluded that the integration between corporate governance and fraud prevention mechanisms, including compliance, auditing, and institutional transparency, is fundamental for the resilience of the financial system. Effective prevention requires the union of technology, governance, and organizational culture to strengthen user security and trust.

Keywords: Business intelligence; Corporate governance; Special Chargeback Mechanism; Pix; Prevention.

November 12, 2026

The potential of Artificial Intelligence in the pricing of Software development

The pricing of software development projects, especially in the context of customizations carried out by development companies and implementation consultancies, represented a recurring challenge due to the subjectivity of estimates and the difficulty in standardizing the criteria used. The study aimed to analyze the application of Artificial Intelligence as a support tool for the software customization pricing process. The adopted methodology was characterized as an exploratory and quantitative study, structured from a comparative experiment. In it, pricing performed by traditional methods was analyzed and these values were compared with estimates generated by Artificial Intelligence models, which allowed identifying divergences, possible biases, and differences in process agility. The investigation used 20 customizations previously estimated by a human expert consultant in OSGT software. The results indicated that the use of Artificial Intelligence provided greater speed in generating estimates and favored the standardization of results, in addition to enabling greater scalability through the simultaneous processing of multiple pricing requests. It was also observed that the effectiveness of the generated estimates depended directly on the quality and level of detail of the scope of the customizations analyzed, concluding that AI is a viable and relevant tool for decision support.

Keywords: Process automation; Software customization; Effort estimation; Estimate standardization.

November 12, 2026

Evaluation of a Backend Architecture Applied to Currency Exchange and Fraud Detection with Artificial Intelligence

The growth of digital financial transactions in recent decades has demanded more efficient validation and fraud prevention mechanisms, requiring scalable and integrated technological solutions. A RESTful API was developed for managing financial transactions, currency conversion based on official exchange rates from the United States Treasury Department, and automated fraud risk analysis. The solution was implemented in Java language, using the Spring Boot framework, structured in a layered architecture, and integrated with external services for obtaining exchange rate data and intelligent transaction processing. The system was divided into modules for transaction registration and consultation, automatic currency conversion, and detection of suspicious behaviors with risk classification. Functional tests demonstrated an average response time of less than 20 milliseconds for internal operations and between 3 and 4 seconds for external processing-dependent anti-fraud analyses. In a simulated scenario, a transaction of US$9,000.00 was correctly classified as high risk due to significant divergence from the analyzed history. The results indicated the technical viability of the solution for applications in fintechs, auditing, and corporate systems, although future improvements could include supervised machine learning models, distributed caching, and asynchronous processing.

Keywords: RESTful API; Currency conversion; Fraud detection; Software engineering; Spring Boot.

November 12, 2026

Company segmentation by clustering reveals distinct financial patterns

The growing availability of corporate financial data has increased the importance of analytical techniques to identify patterns and support decisions in contexts of high heterogeneity among companies. The study aimed to apply clustering techniques to a financial database, comparing grouping methods and statistically validating the results, in order to identify segments with distinct economic-financial characteristics. Methodologically, the research was classified as applied, exploratory, descriptive, and quantitative. Secondary data from companies were used, subjected to inspection, treatment, standardization, and computational modeling. The K-means algorithm and agglomerative hierarchical methods were employed, defining the optimal number of clusters with metrics such as WCSS, Silhouette, Calinski-Harabasz, and Davies-Bouldin. The groupings were evaluated by analysis of variance (ANOVA) and compared by the Adjusted Rand Index (ARI). The results revealed that the solution with six clusters offered the best balance between statistical quality and analytical interpretability, enabling the identification of groups with distinct profiles, such as large corporations, robust intermediate companies, organizations with critical performance, and extreme cases. It was also observed that different algorithms captured distinct structures in the data, highlighting the importance of methodological comparison. It was concluded that clustering constituted an effective approach to segment financial databases and reveal relevant patterns, contributing to more refined analyses and to the applied use of data science in business contexts.

Keywords: Financial analysis; Unsupervised learning; Data science; Clustering; Company segmentation.

November 12, 2026

The Relevance of Emotional Intelligence and “Mindfulness” in Organizations

In the contemporary organizational context, marked by high competitiveness and pressure for results, emotional and psychological aspects of employees have become crucial for performance and well-being. This study analyzed employees’ perception of Emotional Intelligence and “mindfulness” practices in the workplace, investigating their perceived effects on well-being, professional performance, and interpersonal relationships. The research, of a quantitative and descriptive nature, employed the survey method to collect data. A sample of 78 professionals from various organizational sectors responded to a structured questionnaire, with items measured by a Likert scale, based on theoretical models of Emotional Intelligence and “mindfulness”. The data were processed by descriptive statistics using Microsoft Excel software. The results revealed a predominantly positive perception of the participants regarding both constructs, highlighting individual emotional competencies and benefits of mindfulness, such as focus, stress management, and emotional well-being. However, weaknesses were identified in organizational support for the systematic development of these competencies. It was concluded that, despite the individual recognition of Emotional Intelligence and “mindfulness”, there is a demand for greater institutionalization of these practices to optimize their effects in the corporate environment.

Keywords: Mindfulness; Well-being at work; Organizational climate; Socio-emotional competencies; People management.

November 12, 2026

Air pollution and hospital admissions for respiratory diseases in São Paulo (2015–2024)

The relationship between air pollution and hospital admissions for respiratory diseases in the municipality of São Paulo, from January 2015 to December 2024, was investigated. The objective was to analyze the association between the monthly averages of atmospheric pollutant concentrations (PM2.5, PM10, and NO2) and the monthly totals of hospital admissions for respiratory diseases (ICD-10: J00-J99). An observational, ecological, and quantitative study was conducted, based on monthly time series. Data were obtained from public databases of CETESB (QUALAR), DATASUS (SIH/SUS), and INMET, and processed in a Python/Jupyter Notebook environment. Generalized Linear Models were used, specifically the Negative Binomial model, due to the overdispersion identified in the data. The models controlled for seasonality, temporal trend, meteorological variables (temperature and relative humidity), and the COVID-19 pandemic period. No statistically significant association was identified between the pollutants and monthly respiratory admissions in the adjusted models. Sensitivity analysis showed that the initially observed positive association was explained by shared seasonal patterns between pollution and admissions, becoming null after adequate control. It was concluded that, on a monthly time scale, seasonality was the main determinant of the variability in respiratory admissions, limiting the detection of independent effects of air pollution. This result does not contradict the literature but reflects the limitation of monthly aggregation in capturing acute effects. The developed computational pipeline represents a methodological contribution for future investigations with higher granularity data.

Keywords: Respiratory admissions; Air pollution; Air quality; Seasonality; Time series.