Floor Plan-Agnostic Detection of Gait Speed Drifts Using Ambient Sensors
1 st Marina Vicini, 2 nd Martin Rudorfer, 3 rd Zhuangzhuang Dai, 4 th Ahmad Beltagui, 5 th Luis J. Manso · 2026
Read Floor Plan-Agnostic Detection of Gait Speed Drifts Using Ambient Sensors by 1 st Marina Vicini, 2 nd Martin Rudorfer, 3 rd Zhuangzhuang Dai, 4 th Ahmad Beltagui, 5 th Luis J. Manso (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-2605-00900.txt.
Introduction
Marina Vicini, Martin Rudorfer, Zhuangzhuang Dai, Ahmad Beltagui, and Luis J. Manso present a method to detect drifts in gait speed at home without needing a floor plan. Gait speed is a strong health signal for older adults, and ambient sensors can monitor it while preserving privacy. Many earlier methods require the home layout, which is often unavailable, limiting real use.
The approach finds informative transitions between sparse sensors and tracks changes in how long those transitions take. A non-parametric test compares a recent period with an initial baseline for each sequence, and daily results are combined for a robust signal. On simulated data across four layouts, performance matches or beats a baseline that does use floor plan information.
Catalog record
| CREATOR | 1 st Marina Vicini, 2 nd Martin Rudorfer, 3 rd Zhuangzhuang Dai, 4 th Ahmad Beltagui, 5 th Luis J. Manso |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 30 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.00900
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Floor Plan-Agnostic Detection of Gait Speed Drifts Using Ambient Sensors Thanks: This study is conducted as part of Innovate UK Knowledge Transfer Partnership project reference 10062173. The financial support from Legrand Care and Innovate UK is gratefully acknowledged. 1 st Marina Vicini, 2 nd Martin Rudorfer, 3 rd Zhuangzhuang Dai, 4 th Ahmad Beltagui, 5 th Luis J. Manso Abstract Gait speed is a vital health indicator for older adults, as changes in gait speed can reflect physiological and functional decline. Ambient sensors offer a promising, privacy-preserving solution for continuous in-home monitoring of gait speed; although it is often limited by methods requiring a home floor plan, which is frequently unfeasible. This paper proposes a novel, floor plan-agnostic method to detect gait speed drifts using only sparse ambient sensors. Our approach identifies informative sensor-to-sensor transitions and analyses fluctuations in their duration. For each sequence a non-parametric statistical test detects changes between a recent period and an initial baseline; and daily test results are aggregated to provide a robust drift detection response. We evaluate our method on a simulated dataset across four different home layouts, showing performance comparable to, and in some cases exceeding, a state-of-the-art baseline that requires floor plan information. This work demonstrates a feasible approach for scalable, cost effective gait drift detection monitoring, providing a foundation for future validation in complex real-world environments. (a) Layout A (b) Layout B (c) Layout C (d) Layout D Fig. 1 : The four simulated studio apartment layouts used in the experiments, generated using the simulator from [ 32 , 31 ] . The layouts feature different furniture arrangements and sparse sensor placements representative of commercial installations. I Introduction For the ageing population, mobility is a key performance indicator of health. Gradual physical decline places older adults at risk of losing their independence and potentially needing to relocate [ 29 ] , which can compromise their right to self-determination. Notably, regularly monitoring the health condition of older adults is crucial for detecting early-stage functional decline, as this decline shows a considerable individual variability [ 25 ] ; it allows proactive and preventative care and making ageing-in-place a reality for longer. Gait is a complex activity that integrates the nervous, musculoskeletal, and cardiovascular systems [ 36 ] , and a reliable measure for assessing the functional status of older adults [ 21 ] . A decline in gait speed is a consistent risk factor for adverse outcomes such as hospitalisation, mortality, cognitive impairment, and disability [ 2 ] , and can reflect subtle yet significant functional decline before more visible symptoms manifest. A change in gait speed of 0.10\;m/s is widely considered a minimal clinically important difference [ 8 ] . Traditionally, gait speed is assessed through tests in clinical settings, such as standardised timed walks ( e.g. , a 4-metre walk) [ 13 ] . While valuable, these clinical assessments must be performed under the supervision of a professional, such as a clinician or physiotherapist. A significant shortcoming is their infrequent administration, providing only a snapshot of a person’s health. Long-term monitoring can capture variations and gradual changes, and therefore, it has the potential for early diagnosis of various diseases and fall risk assessment, allowing for more timely interventions [ 22 ] . Home-based assessment allows for this continuous monitoring. Ambient Intelligence (AmI) and its application in healthcare, Ambient Assisted Living (AAL), leverage sensorised environments that facilitate continuous monitoring. A variety of sensor technologies can be used for gait analysis. Wearable sensors, such as pressure-sensing insoles [ 7 ] , Inertial Measurement Units (IMUs) [ 28 ] , [ 20 ] , or smartphones [ 18 ] , can provide detailed gait metrics but require users to consistently wear and charge the devices, which is not suitable for all older adults. In contrast, ambient sensors do not require user interaction. This category includes technologies like Ultra-Wideband (UWB) or mmWave radar [ 17 , 1 ] , cameras [ 9 ] , radio signals [ 16 ] , and Wi-Fi sensing [ 37 ] . However, the cost, privacy implications, or installation complexity of these systems can be barriers to widespread commercial adoption. Motion detectors, such as Passive Infrared (PIR) sensors, and door contact sensors are less data rich but offer a cost-effective and privacy-preserving alternative for mobility assessment [ 30 ] . The motivation to using these sensors is enhanced by the proliferation of commercially available smart home kits and telecare services, such as those promoted by advisory body like the Technology Enabled Care Services Association in the UK [ 4 ] , that indicate that over 2 million people in the UK use technology enabled care (TEC). These general-purpose systems are already being deployed to support older adults, typically relying on a core set of cost-effective ambient sensors like PIR motion detectors, door contacts, and smart plugs [ 10 , 3 ] . Their primary objectives are broad, including abnormal behaviour detection, medication prompting, and general Activities of Daily Living (ADL) recognition [ 35 , 33 , 12 ] , rather than specialised gait analysis. Integrating gait speed drift detection capabilities into already existing systems creates a significant opportunity, as this vital health metric could be monitored at scale without requiring any new, specialised hardware. However, the existing literature for gait speed drift detection in smart homes requires either a detailed floor plan of the home or a specific configuration of sensors in a line [ 14 , 11 ] . Obtaining a floor plan is not economically feasible, and creating specific setups would drive the cost for both sensors and installation. These observations lead to our central research question: Is it possible to detect gait drifts using ambient sensors without a-priori knowledge of the home’s floor plan? Unlocking this capability would create a huge impact for existing and already deployed commercial systems. This paper proposes a method to detect gait drifts in a smart home environment using only ambient motion and door sensors, without relying on a floor plan. The contributions of this paper are as follows: (1) a problem formulation for gait speed drift detection using sparse ambient sensor data; (2) a simulated dataset containing 4 different house layouts and a wide range of gait speed options; and (3) a method to detect gait speed drifts in a smart home without relying on any additional spatial relation between the sensors. II Related Works To the best of our knowledge, there are no methods that aim to detect changes in gait speed from ambient sensors, without requiring any spatial context. However, there are several methods that use spatial context to directly estimate gait speed, which are most closely related to our work. These can be categorised into: (1) motion sensors placed in a line; and (2) sensors distributed across the house. II-A Sensor Array A common approach involves distributing an array of PIR sensors in a line, typically along a hallway ceiling or a bespoke rail, to measure the time between sensor activations. Knowing the fixed distance between sensors allows for a direct calculation of walking speed. Hayes et al. [ 15 ] pioneered this method by positing four motion sensors with a restricted field of view along a path