Non-invasive techniques for flow rate measurement in water pipes: Systematic review

Non-invasive flow measurement in water distribution systems is imperative for ensuring water efficiency, leak
detection, and sustainable resource management. Recent technological advances have introduced alternative
solutions that rely on external sensing and signal processing rather than intrusive installation. However, there
is still a lack of consolidated evidence regarding their accuracy, energy efficiency, and applicability across
different contexts. In this paper, we present a systematic review conducted under a Kitchenham-style protocol,
covering 32 primary studies focused on non-invasive techniques for flow monitoring in pipes. The analysis
indicates that ultrasonic clamp-on devices consistently attain measurement errors below 2% in controlled
environments, while accelerometer-based solutions demonstrate reduced precision (5%–10% error) yet excel
in low-cost and retrofit scenarios. The employment of signal processing methodologies, encompassing time-offlight analysis, spectral decomposition, and cross-correlation, has been demonstrated to markedly enhance the
detection capabilities of the system. Moreover, the integration of Machine Learning (ML) algorithms has been
shown to enhance the classification accuracy by up to 15% in comparison with conventional threshold-based
approaches. However, less than 20% of the studies provided detailed energy consumption metrics, and only
a minority validated their systems in real-world installations. These findings underscore two salient points.
Firstly, they highlight the maturity of ultrasonic methods. Secondly, they underscore the untapped potential
of accelerometer-based solutions. The findings also underscore the need for standardized evaluation protocols,
energy-aware designs, and large-scale field validations. Such measures are necessary to bridge the gap between
laboratory prototypes and operational deployments.