Routine Computing: A Systematic Review of Sensing Daily Life Dimensions Towards Human-Centered Goals Conference: Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems; April 13–17, 2026; Barcelona, Spain Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems (CHI ’26), April 13–17, 2026, Barcelona, Spain DOI: 10.1145/3772318.3791269 ISBN: 979-8-4007-2278-3/2026/04 CCS: Human-centered computing Empirical studies in HCI CCS: General and reference Surveys and overviews

Borislav Pavlov, Jiajin Li, Jun Fang, Yuntao Wang, Yuanchun Shi · 2026

Read Routine Computing: A Systematic Review of Sensing Daily Life Dimensions Towards Human-Centered Goals Conference: Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems; April 13–17, 2026; Barcelona, Spain Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems (CHI ’26), April 13–17, 2026, Barcelona, Spain DOI: 10.1145/3772318.3791269 ISBN: 979-8-4007-2278-3/2026/04 CCS: Human-centered computing Empirical studies in HCI CCS: General and reference Surveys and overviews by Borislav Pavlov, Jiajin Li, Jun Fang, Yuntao Wang, Yuanchun Shi (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-21934.txt.

Introduction

Pavlov and four colleagues present the first review of routine computing, a field focused on sensing and modelling human behaviours. The review covers 203 studies published up to August 2025 and presents a new map of time patterns, behaviour links, thinking aspects, and change and drift. Common goals span four domains: care access, healthy habits, adaptive and context-aware support, and group insights. Hard problems include the gap between low-level activity spotting and high-level intent, personal versus general tension, privacy worries, and data limits.

Catalog record

CREATORBorislav Pavlov, Jiajin Li, Jun Fang, Yuntao Wang, Yuanchun Shi
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE78229
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.21934

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Routine Computing: A Systematic Review of Sensing Daily Life Dimensions Towards Human-Centered Goals Conference: Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems; April 13–17, 2026; Barcelona, Spain Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems (CHI ’26), April 13–17, 2026, Barcelona, Spain DOI: 10.1145/3772318.3791269 ISBN: 979-8-4007-2278-3/2026/04 CCS: Human-centered computing Empirical studies in HCI CCS: General and reference Surveys and overviews Borislav Pavlov, Jiajin Li, Jun Fang, Yuntao Wang, Yuanchun Shi Abstract . Human routines structure daily life, yet remain challenging for computational systems to understand. This paper presents the first systematic review of routine computing, a previously implicit but increasingly recognized field that focuses on computationally sensing and modeling human behaviors. It synthesizes 203 studies published up to August 2025. The paper presents a new taxonomy of the literature, focusing on temporal structures, behavioral interactions, cognitive aspects, and how variability and deviations are addressed. The common goals of routine computing extend across four major application domains, including accessibility care, the promotion of healthy habits, adaptive and context-aware support, and large-scale population insights. Persistent challenges that limit the design of truly human-centered systems are identified, including the gap between low-level activity recognition and high-level intent, the tension between personalization and generalization, unresolved privacy concerns, and data-related limitations. By consolidating these findings, this paper provides a foundational framework for HCI researchers, outlining principles for designing ethical, adaptive, and human-centered routine-aware systems. † † cc-license: by 1. Introduction Figure 1. Visual abstract of routine computing survey and taxonomy. This framework summarizes the conceptual landscape of routine computing across three major layers. Dimensions outline four core perspectives for characterizing everyday routines: Temporal Rhythm , Behavior Context , Cognitive Factors , and Adaptation Principles . Goals describe four primary application targets, including Aging and Disability Care , Promoting Healthy Habits , Adaptive Support , and Population Insights . Future Directions highlight emerging opportunities for advancing the field through Understanding Human Intention , Personalized Adaptation , Design for Privacy and Ethics , and Enriching the Data Chasm . Diagram showing the dimensions of routine computing: sensing granularity, behavioral and social contexts, cognitive and psychological factors, and application domains. Human life is structured by routines—recurring sequences of activities such as commuting, eating, working, and resting—that give rhythm, meaning, and stability to everyday experience. Recent advancements in ubiquitous computing have sought to integrate more seamlessly into these daily patterns by modeling and responding to human behavior. Yet, as Suchman’s Situated Action ( 184 ) and Dourish’s Embodied Interaction ( 49 ) remind us, human behavior is deeply contextual and interpretive, making the computational modeling of routines particularly demanding. The study of human routines in computing draws on several long-standing research traditions. Human Activity Recognition (HAR) research, beginning with foundational work by Katz ( 99 ) and Lawton and Brody ( 115 ) , established formal taxonomies of self-care and instrumental tasks that later informed monitoring practices in gerontology, disability care, and home-based health assessment. Context-aware computing ( 1 ) extended this perspective by emphasizing that meaningful interaction requires sensing and interpreting the situational factors surrounding an activity—location, timing, social presence, and environmental state. Across these traditions, prior surveys have offered important overviews of related areas ( 14 ; 169 ) , however, they were not designed to address the broader patterns that structure daily life. They typically focus on momentary or single-category activities, rather than the complex, evolving, and interdependent actions that unfold across hours, days, and weeks. To address this gap, we introduce the concept of routine computing , defined as the computational sensing, modeling, understanding, and application of recurring human behaviors in daily life. Routine computing extends beyond activity recognition by foregrounding the temporal, contextual, and psychological dimensions that give everyday practices their meaning and significance. Our work is guided by the following research questions: RQ1: What conceptual dimensions are emphasized by existing studies of daily routines in HCI and ubiquitous computing research? RQ2: What goals drive the development of routine-aware technologies? RQ3: How do the identified dimensions and goals of routine computing shape the user-experience challenges for designing future human-centered routine-aware systems? These research questions were formulated to clarify the foundational structure of routine computing as a research area. RQ1 aims to identify the underlying dimensions that organize existing approaches and to establish a shared conceptual vocabulary. At the same time, routine-aware systems are developed with diverse aims—from supporting health and aging to improving productivity or understanding community behavior—yet these motivations have not been systematically compared. RQ2 addresses this by examining the functional goals that shape the development of routine-oriented technologies and the problems they aim to solve. Finally, as routine computing increasingly influences user-facing technologies, it raises questions about how such systems should interact with people, integrate into everyday life, and support long-term engagement. RQ3 therefore explores the design and user-experience considerations that follow from routine-aware computation. Together, these questions provide a structured basis for synthesizing the literature and defining the scope of routine computing as an emerging research domain. Therefore, in this paper, we conduct a systematic review of routine computing in HCI and ubiquitous computing, synthesizing 203 studies up to August 2025. Our analysis classifies work by sensing temporal granularity, behavioral and social contexts, cognitive and psychological factors, and adaptation strategies. The review examines routine-based applications in domains such as aging, health, self-reflection, and population-level insights. Accordingly, this paper makes the following contributions: (1) We provide the first systematic review of routine computing in daily life, synthesizing 203 studies across sensing, modeling, and applied domains. (2) We articulate routine computing as a conceptual foundation , distinguishing it from related areas such as activity recognition and context-aware computing. (3) We develop a taxonomy of routine computation along temporal, behavioral, cognitive, and adaptive dimensions, consolidating diverse methods and approaches. (4) We identify critical challenges and future opportunities , including interpretability, personalization, ethical design, and data enrichment, setting a research agenda for advancing routine-aware HCI. 2. Methodology PRISMA flowchart detailing the article selection procedure. The initial search identified 2,631 records from two primary databases: ACM DL (574 records) and IEEE Xplore (2,057 records). After screening for duplicates, 10 were excluded, leaving 2,621 records for title, abstract, and outline review. During this screening, 2,225 records were excluded based on criteria such as not being written in English, non-human subjects, or focusing on technological system routines. A full-text assessment was then performed on the remaining 396 records. The final eligibility screen excluded 193 more records if they were not primary research, not relevant to recurring or daily-life behavior patterns, or presented redundant solutions. Ultimately, 203 records were included in the final review, with 77 papers from ACM DL and 126 from IEEE Xplore. Figure 2. PRISMA flowchart detailing the article selection procedure PRISMA flowchart detailing the article selection procedure. The initial search identified 2,631 records from two primary databases: ACM DL (574 records) and IEEE Xplore (2,057 records). After screening for duplicates, 10 were excluded, leaving 2,621 records for title, abstract, and outline review. During this screening, 2,225 records were excluded based on criteria such as not being written in English, non-human subjects, or focusing on technological system routines. A full-text assessment was then performed on the remaining 396

Read the complete text