Sensors have recently become valuable tools in engineering, providing real-time data for monitoring structures and the environment. They are also emerging as new tools in education and training, offering learners real-time information to reinforce their understanding of engineering concepts. However, sensing technology’s complexity, costs, fabrication and implementation challenges often hinder engineers’ exploration. Simplifying these aspects could make sensors more accessible to engineering students. In this study, the researcher developed, fabricated, and tested an efficient low-cost wireless intelligent sensor aimed at education and research, named LEWIS1. This paper describes the hardware and software architecture of the first prototype and their use, as well as the proposed new versions, LEWIS1-β and LEWIS1-γ, which simplify both hardware and software. The capabilities of the proposed sensor are compared with those of an accurate commercial PCB sensor. This paper also demonstrates examples of outreach efforts and suggests the adoption of the newer versions of LEWIS1 as tools for education and research. The authors also investigated the number of activities and sensor-building workshops that have been conducted since 2015 using the LEWIS sensor, showing an increasing trend in the excitement of people from various professions to participate and learn sensor fabrication.

Rebar spacing inspection is a critical, mainly manual quality control task in reinforced concrete construction. Existing augmented reality (AR) inspection tools are unidirectional: information flows from machine to inspector, who cannot feed field observations back into the analysis. This study presents a closed-loop AR interface for human–machine teaming in rebar inspection, integrating RGBD sensing, density-histogram detection with principal component analysis line fitting, greedy correction optimization, and a HoloLens 2 interface over bidirectional TCP/IP. Three loops carry information from human to machine: manual measurements that override machine spacings and trigger re-optimization, per-step confirm/reject decisions that filter corrections through construction knowledge, and inspector-initiated rescans that verify physical corrections on fresh data within the same session. The paper introduces a complexity decomposition separating observable complexity (Kobs), which reflects sensor-perceived geometric correction difficulty, from hidden complexity (Khid), which represents constraints invisible to the sensing system. Across five controlled configurations, single-pass correction resolved all violations in the tested cases with Khid=0, whereas both configurations containing a sensor-invisible constraint left residual violations. The bidirectional loop reached zero residual violations in all five configurations by allowing the inspector to communicate physical constraints unavailable to the sensing system. Validation also demonstrated sub-millimeter agreement with caliper ground truth and scan-to-scan repeatability. These results provide a proof-of-concept demonstration of closed-loop AR inspection and illustrate how sensor-invisible feasibility constraints can create failure modes that cannot be resolved from sensing alone.

Accurate rebar placement is essential for ensuring the structural performance and durability of reinforced concrete components, particularly in construction industry. In practice, rebar inspection is commonly performed using manual measurements and two-dimensional drawings, resulting in subjective interpretation, inefficiencies, and physically demanding inspection workflows. This study presents an augmented reality (AR)–based rebar inspection and correction system developed to support practical inspection and decision-making tasks. The system integrates RGBD sensing, automated Building Information Modeling (BIM)–based spacing analysis, greedy correction sequencing, and motion-capture-based ergonomic assessment within a unified workflow. As-built rebar geometry is captured using an Azure Kinect depth camera and registered with the design BIM model to identify spacing deviations. Inspection results are presented in situ through a Microsoft HoloLens 2 headset using spatially anchored, color-coded AR overlays, while a greedy optimization algorithm generates step-by-step correction guidance intended to minimize unnecessary bar adjustments and physical effort during inspection. Real-time TCP/IP communication enables low-latency transfer of analytical results to the AR interface, supporting interactive inspection without offline processing. A controlled laboratory study compares conventional manual inspection procedures with AR-assisted workflows using a Vicon motion capture system to quantify differences in task execution and inspector movement. The results demonstrated descriptive evidence of differences in task duration, movement patterns, direction changes, and trunk and neck postures between inspection approaches. The findings demonstrate the practical potential of an integrated digital workflow to enhance quality control in construction, addressing industry-identified needs for objective measurement, reduced physical effort, and guided decision-making in repetitive inspection tasks.