Abstract

The safety of workers is significant when talking about large risky workplaces such as construction job sites. Therefore, we proposed a UHF-RFID tag antenna that can be used to identify various Job site equipments and also construction workers by tagging their wearable helmets. Here, the compact, low profile, and simple structured platform-tolerant RFID tag antenna is proposed and optimized for metal sheet and ABS-based safety helmets. Furthermore, WTMiddleware (Worker Tracking Middleware) using Javascript, CSS, and HTML has been implemented to demonstrate a real-time secured web-based remote access of workers to provide their actual on-site monitoring and tracking of construction equipment carried/operated by the workers for tracking solutions. First, the fabricated tag’s performance was tested successfully in an outdoor scenario on various mountable materials such as wood, glass, plastic, and metal. The tag yielded a reasonable read range of 5.9 m on metallic sheet and 7.1 m on plastic with the highest read distance of 9.3 m on a glass sheet. Also, the measurement setup along with developed WTMiddleware was deployed in an actual workplace to demonstrate practical utility of the designed tag affixed on workers’ safety helmets. It was observed that all the workers wearing tagged helmets were successfully identified, tracked, and monitored. Also, their real-time on-site details, working zone, and complete personal details were accessed through a developed web page application via WTMiddleware.

1. Introduction

In large job sites, constant supervision is required for site workers’ safety besides taking care of different equipment used during construction. Additionally, managing effective access control in construction job sites poses a significant challenge in terms of the complexity of risky or construction site’s layout [1, 2]. Therefore, it is necessary to monitor the workers and other staff for their safety concerns and effective site management. Hence, a traceability system is considered one of the necessary factors in tracking and tracing these workers and equipment at riskier workplaces.

Recently, Internet of Things (IoT)-enabled radio frequency identification (RFID) systems have been recently utilized in traceability systems, such as for food [3], material management [4], smart logistic applications [5], warehouse management [6], and many more. RFID, an auto-identification technology, is a key enabling technology for implementing an IoT network [7]. The data of the tagged object is stored in the RFID chip integrated with the tag. The reader captures this stored data, and the acquired information is further transferred to the cloud database for remote access and monitoring objects. Hence, RFID is not only being employed for identification but also for special applications, e.g., hydrogen concentration monitoring [8], structural health monitoring [9], screw relaxing monitoring [10], and human parameters monitoring [11]. This is possible due to its benefits of easy deployment and low-cost solutions [1214]. IoT-enabled RFID systems are also pivotal for achieving accurate tracking and effective access control at construction sites. In [15], the author discussed the possibility of developing integrated smart OHS (occupational health and safety) system design for construction sites by implementing monitoring-enabled wearable equipment and making them mandatory for the workers. But the implementation of a complete IoT-enabled RFID system with backend server connectivity for construction sites has never been reported in the literature.

The classic UHF-RFID system comprises a reader device equipped with reader antennas and tags. The material of the objects to which the tags are affixed significantly interferes with the RFID operation. Therefore, it is essential to evaluate the performance of RFID tags [16] under the influence of equipped object material. In developing a UHF-RFID system, tag antenna design faces a significant challenge due to interference from surrounding objects or environments [17]. The noises generated due to the multipath effect and signal interference deteriorates the backscattering process. Also, as the tag needs to be placed on different objects to be tracked, its radiation performance is strongly affected [18]. Due to the presence of conductive objects such as metal and liquid-containing objects, the incident signal is added destructively to the image current from these surfaces [19]. This results in degrading the performance of the tag antenna parameters such as input impedance, power transfer coefficient, radiation efficiency, and realized gain. More commonly, the passive tag must be mounted at a distance from the metal surface (around 1 cm), as stated in [20]. Also, there should be special considerations if some other nearby electromagnetic sources is working with the same operating frequency range. A more comprehensive study has been represented in [21], which discusses the interference of metal surfaces on passive tags. The author demonstrated that the tag antenna receives more power to wake up the chip when placed at λ/4 (being λ the wavelength) distance from the surface. Different solutions are proposed in the recent literature to overcome these constraints for improving the performance of the tag antenna [2225]. Recently, platform-tolerant tags have been designed to exhibit a higher read range in free space and when mounted on metal surfaces. There are commercial tags available that are mountable on metal. However, they are having larger dimensions with a bulky structure [26, 27].

Furthermore, a Middleware must be developed with RFID technology to implement the entire IoT-enabled UHF-RFID RFID traceability system. RFID middleware acts as an interface between the RFID hardware system and its application. Thus, middleware handles all the incoming real-time data from the connected reader without any read miss. Furthermore, it also serves to process the counting, filtering, and collecting of the tag data [2832]. In [33], Haibi et al. developed a Middleware named BTMiddleware that uses MongoDB NoSQL database. This middleware was implemented for baggage traceability at airports, and the program was written in NetBeans IDE as the development environment using Java as the backend language. The user has the manual task of adding the reader name and IP address of the reader on the home page of the middleware to access the information. Remote access to the data recorded has been provided. Also, in [5], a Middleware combining block chain and RFID was designed to make the supply chain robust and secure. In this work, liquid bottles were tracked as they moved forward in the supply chain from the manufacturer to the retailer by attaching tags to them. The mobile application for remote access was implemented in a real-world scenario using block chain technology. It has a high accuracy varying between 92.5 and 97.5 percent. However, the time for mining each block varies due to the difference in input string lengths resulting in the variation of the transaction cost, which is the drawback of block chain. Extending the given concept for a risky construction site, we have designed an IoT-enabled UHF-RFID system by developing a worker tracking WTMiddleware. The proposed approach demonstrates a real-time secure web-based solution for workers’ tracking, actual site monitoring, and tracking of vehicles/equipment. Following contributions are made in this study:(i)The compact microstrip patch dipole configuration-based UHF-RFID tag antenna has been designed and optimized for mounting hard-to-tag metal-based equipment and the safety helmet.(ii)The read range performance of the proposed RFID tag design has been experimentally validated on a metallic sheet and safety helmet using Impinj R420 reader and circularly polarized (CP) RFID reader antenna.(iii)We designed WTMiddleware for enabling traceability and effective access control using a cloud-hosted NoSQL database. The developed middleware shows the potential to revolutionize effective management at the job site.

The robustness and feasibility of the proposed IoT-enabled UHF-RFID tracking system is demonstrated by experimental testing at a real job site scenario. The rest of the article is organized as follows: Section II, “IoT-Enabled UHF-RFID system architecture,” introduces the analogous hardware-hardware configuration. After that, section III, “Integrated RFID system hardware with WTMiddleware,” provides the relevant description of hardware and software development features. This section also provides guidelines regarding the design in the subsection titled “Tag Design and Analysis.” The optimization of the tag antenna design for mounting it on a metal sheet and ABS plastic-based safety helmet is shown in the subsection titled “Parametric Study of proposed dipole tag.”

Furthermore, the impedance characterization, read range measurements, and their validation for the fabricated prototype of the designed tag is presented. The software implementation and hardware management interface details are given under the subsection “WTMiddleware Design.” Furthermore, the details of on-site testing of the complete system are outlined under the “Real-time testing” section. Finally, the significant contributions of this research study are discussed under the “Conclusions” section.

2. IoT Enabled UHF-RFID System Architecture

The system model to attain total digital worksite visibility of workers using the IoT-enabled RFID system is illustrated in Figure 1. The proposed scheme provides real-time application for tracking and providing access control at the job site. The proposed framework has the following significant blocks:

2.1. Personnel/Equipment Reading Block

The reader transmits a continuous wave (CW) signal for a specific duration (predefined) with maximum power to charge all the tags under its coverage range. Thus, in the personnel/equipment reading block, the data related to worker/equipment (hauling/lifting) will be read from the RFID affixed gadget and registered in the database.

2.2. Data Acquisition Block

In the data acquisition block, the tagged gadget/equipment data will be read by deployed UHF reader and fed to the WTMiddleware. Here, the reader identifies the tagged workers falling in the range of the reader.

2.3. WTMiddleware Block

In the WTMiddleware block, acquired data is further processed, translated, and authenticated using a web-based application for remote access by the supervisor. Here, an onsite monitoring operation is performed to verify workers present at the job site. If any worker is not present or the tag is not replying, it will inform the monitoring computer to raise the alarm to notify the supervisor. A detailed explanation of the integrated hardware-Middleware is given in the next section.

3. Integrated RFID System Hardware with WTMiddleware

RFID system hardware consists of an RFID reader connected to the reader antenna, RFID tag, and host computer. For the WTMiddleware design, CSS and HTML have been used to design the webpage frontend; NoSQL and Javascript have been used as backend language. The typical view of the integration of RFID system hardware with WTMiddleware is shown in Figure 2. A detailed discussion of hardware and WTMiddleware is given below.

3.1. RFID System Hardware
3.1.1. RFID Reader

In the proposed system, we used a 4-port Impinj R420 reader with 32-read zones to support different locations around the working or job site area in a complex construction site. The deployed reader has a receiver sensitivity of −84 dBm with a maximum transmit power of 32.5 dBm. The reader is connected to a circularly polarized broadband (860–940 MHz) reader antenna. The reader interrogates the tagged objects by sending the modulated signal towards the tag. The reader receives the backscattered signal from the tag containing the required information. The installed reader will automatically receive the data transmitted by the tag-equipped gadgets from different locations.

3.1.2. RFID Tag

The details of each worker, including their entry/exit time and location, are stored in the RFID tag mounted on the safety helmet. The function of the tag is to transmit the worker’s unique identity (ID) number to the allocated RFID reader. Thus, each worker under the specific reading zone is continuously scanned and monitored in real-time. Additionally, tracking metal-based equipment will efficiently manage the overall construction site. This will also prevent mishandling of harmful equipment.

(1) Tag Design and Analysis. As already discussed, the tag performance gets detuned when tagged on different objects such as metal and objects with different dielectric properties (liquid-containing bottles, plastic, wood, glass, etc.). Therefore, the codesign approach was implemented, considering the effect of the metallic object and ABS-based helmets, respectively. Here, a microstrip patch-based dipole tag antenna is designed to work in the ETSI band (865–867 MHz) to provide a low-cost, platform-tolerant, and most optimal solution. The proposed UHF-RFID tag and codesign model (on metal and safety helmet) are shown in Figures 3 and 4, respectively. The proposed tag comprises a two-layered structure with the top layer as a radiating dipole and the bottom layer as a metallic ground.

Two shorting pins/vias are employed at its edges to connect the top patch layer and the ground plane for the design purpose. The Alien Higgs-3 RFID chip [34] is assumed to be integrated into the middle of the dipole arms of the antenna configuration. The impedance of the employed Higgs-3 chip was obtained considering its parameters Cic = 0.85 pF and Ric = 1.5 k provided in the datasheet [34] with an equivalent impedance value of 27-j 212 j at 866 MHz. This forms a “3D inductive loop” to exhibit high inductive reactance to compensate for the capacitance of the RFID chip. The tag is optimized by placing it on the metal sheet and ABS-based plastic safety helmets.

(2) Parametric Studies of Proposed Dipole Tag. The rigorous analysis of the designed tag antenna on the metal sheet (200 mm × 200 mm) was conducted by performing its parametric study based on its geometrical dimension variations. Figures 5(a) and 5(b) show that the real and imaginary component of the tag’s impedance decreases as the length of the upper slot, i.e., “L1” increases (varied from 4 mm to 6 mm in steps of 0.5 mm). Also, the gain of the tag antenna increases with increasing “L1” under the influence of the back object, as observed from Figure 6. Hence, the optimum value of slot length, “L1” = 5 mm, was chosen to trade-off between both conjugate impedance and gain.

Furthermore, Figures 7(a) and 7(b) show the variations in the real and imaginary components of tag impedance with respect to variation in space between the dipole arms, i.e., “” from 3.5 to 4.5 mm (in steps of 0.5 mm), respectively. There is a slight increase in the values of both impedance components, i.e., resistance and inductance, with respect to the increase in “.” Also, from Figure 8, the gain of the designed tag increases with an increase in the center space between the dipole arms. This is justified because the field in the opposite dipole gets canceled as the distance “” decreases. This further leads to reduced gain. Therefore, the space between the dipole arms is optimized to achieve conjugate impedance matching and gain. The optimum value of “” is selected to be 4.5 mm.

Moreover, the effect of change in the “” (feed width) parameter on the tag’s impedance is shown in Figure 9. It is observed that both real and imaginary components of the tag impedance rise significantly with the increase in value of “” (varied from 2.48 to 3.48 mm in steps of 0.5 mm). Also, Figure 10 indicates that the gain of the designed tag antenna increases with an increase in value of “.” The optimum value chosen for “” is determined to be 2.98 mm.

Here, the variation in its impedance and gain is studied by varying the size of the metal sheet (200 mm × 200 mm, 300 mm × 300 mm, and 400 mm × 400 mm). It is observed that the impedance of the designed tag remains stable with the increase in the size of the metal sheet (Figure 11), validating its platform tolerance capability. However, from Figure 12, the gain of the tag is observed to decrease as the metal sheet size increases.

The designed tag antenna was simulated and designed using CST software version 2018. Table 1 shows the optimized dimensions of the designed tag antenna for mounting on both platforms, i.e., metal sheet and safety helmet.

The simulated tag antenna was fabricated by etching 0.035 mm copper on a RO 4350 B substrate (permittivity of 3.3 and loss tangent of 0.0031), as shown in Figure 13.

For analyzing the fabricated prototype performance for practical utility, impedance measurement and read range testing was performed. The impedance of the tag was measured by employing the tag inlay without an RFID chip, as shown in Figure 14(a). The impedance of the designed tag was measured using a differential probe set up and vector network analyzer E5071C following port extension calibration as detailed in [35]. The input impedance of the proposed tag antenna, was calculated using the following equation:where is the characteristic impedance of coaxial lines. The helmet’s material was chosen to be ABS plastic having a relative permittivity, εr = 3.5. Furthermore, the employed Higgs-3 chip impedance simulated and measured tag antenna’s impedance after mounting on a metal sheet (with dimensions 200 × 200 mm2) are plotted in Figure 14(b), respectively. It is observed that the measured impedance of the tag is 26.3 + j 222.2 Ohms compared to the simulated tag antenna’s impedance of 22.9 + j 215.1 Ohms at a resonant frequency of 866 MHz, validating a good impedance match in the ETSI RFID band, respectively. The slight difference in the measured impedance compared to simulated impedance may be attributed to the fabrication tolerances and changes in the chip impedance. Also, there are accuracy issues concerned with using a differential probe setup for the impedance measurements method of the tag antenna. Furthermore, Figure 15 depicts the S11 results to validate the impedance matching between the proposed tag antenna and microchip. It is observed that the measurement results agree well with the simulated S11 results.

Furthermore, the RF1 and RF2 terminals of the Higgs-3 RFID microchip were soldered to the feeding port terminals of the designed tag antenna to measure its read range and its platform tolerance capability. The proposed tag’s performance was verified by performing the read range measurement method using the Impinj R420 reader setup [36]. The experimental setup consists of an RFID reader connected to a circularly polarized reader antenna (gain of 9 dBi) and a host computer with an installed Impinj item test software setup. The measurement setup works by transmitting a reader input signal towards the tag-mounted helmet. The tag intercepts the part of this incident signal as there is a power loss accompanied in the free space and the cables of the setup. The RFID chip gets activated by utilizing this intercepted received power, and finally, the tag sends back the backscattered signal towards the reader setup. The host computer finally obtains this backscattered signal and receives data.

The read range performance of the proposed tag is validated in a real outdoor scenario by pasting the prototyped tag on metal and on the helmet and different dielectric surfaces (foam, plastic, wood, and glass), as depicted in Figures 16(a) and 16(b).

The read range was measured by moving the tag away from the reader and determining the maximum distance it could be recognized. Although the proposed tag was designed, simulated, and optimized to mount on metal equipment and wearable helmet, it was observed that it exhibited satisfactory performance on other objects (plastic, wood, foam, and glass) with different dielectric properties. Figure 17 shows the read range performance of the tag on different mounted objects.

The tag exhibited a read range of 5.9 m when pasted on the metal sheet (200 × 200 mm2) and 7.1 m when mounted on an ABS plastic-based safety helmet. Also, the designed tag antenna exhibited a read range of 7 m, 4.9 m, 6.1 m, and 9.3 m when mounted on different material objects such as the plastic sheet, foam, wood, and glass, respectively, validating the designed tag’s robustness.

Figure 18 shows the 3D radiation pattern of the tag at a center frequency of 866 MHz on the metal sheet and safety helmet. It is depicted that the tag pasted on the metal sheet exhibited a directional pattern with a gain of −4.88 dB on the metal sheet (200 × 200 mm2) and −2.1 dB on the helmet (circumference of 10 cm) directivity of 4.15 dBi, demonstrating platform tolerance capability.

The comparison of the proposed tag’s performance with respect to other recently developed tags has been carried out in Table 2.

The proposed tag is observed to exhibit a longer read range than that of tags designed in [3745]. In addition, the size of the proposed tag is smaller than the designed tags in [37, 41, 4347]. On the other hand, the tag’s read range in [40, 47] is higher but with a larger size. In addition, the gain of the proposed tag is more as compared to tags designed in [37, 38, 43, 47], resulting in a higher read range. Furthermore, the tag designed in [46] needs to be placed at a gap of 0.2 mm from the conductive surface. Thus, the proposed tag is found to be potential candidate suitable for mounting on metal, plastic, wood, foam, and glass objects. Also, it has better performance in terms of size, gain, and read range, which is essential for tagging different jobsite equipment and worker’s wearable helmet for tracking and access control. Furthermore, to interrogate the developed RFID prototype from a remote location, a suitable WTMiddleware has been designed for anywhere everywhere access of workers and tagged equipments.

3.1.3. WTMiddleware

The designed tag allocates a unique identity to the workers or equipment, which is necessary to monitor workers in real-time. The web application was deployed to enable tracking of workers by developing a WTMiddleware. In this middleware, a web-based application program is written to display the on-site working details of the workers performing the following main functions:(i)Hiding all the information fetched by the hardware from the backend applications.(ii)Removing all unwanted information that is not required to be shown to the end-user.(iii)Processing the raw data before using and sending it to other applications. It thus provides selectivity in choosing the readers and using information from different readers separately.

Figure 19 shows the WTMiddleware management interface presenting the communication between the reader and a web-based application. Here, CSS and HTML language were used to develop frontend web page design. The backend code for the developed middleware was performed using Javascript.

Reader interface provides an API to integrate data acquired through reader’s item sense software into the website. The readers are connected through Wi-Fi/Ethernet connection to communicate with the installed servers. Furthermore, the middleware management layer also provides link to the database storage using the NoSQL cloud Firebase [28].

In the WTMiddleware design, Algorithm 1 illustrates the authentication of the supervisor to prevent data theft and verify only authorized personnel to access this information.

If the email id and password match the one stored at the backend, the user is taken to a select page with three options: worker details, current on-site details, and logout. The worker details page will have the stored data such as name, address, and contact details and verify the authorized workers from the firebase database. The current on-site details are the page for monitoring on-site worker’s status and location. Finally, the logout page will direct the user to the login page. Figure 20 shows “Login Interface for Supervisor Login and Start-up Interface,” in which the login will be provided to the supervisor through the login interface for authentication purposes.

The authenticated supervisor is required to enter an email id and password for login. In the start-up interface, the webpage will direct the supervisor to observe the information regarding the authorized workers (EPC ID, name, address, and contact) and their real-time onsite details. The pseudocode of Algorithm 1 and the developed page using the algorithm is given as follows:

(1)enter email and password
(2)if (email and password match) then
(3)  go to Select page
(4)  if (selected = worker details) then
(5)    go to page third1
(6)  else if (selected = current on-site details) then
(7)    go to page third2
(8)  else if (selected = logout) then
(9)    go to page index
(10)  End
(11)else
(12)  nvalid id or password
(13)  stay on index page
(14)End

Algorithm 2 demonstrates the process of implementation of on-site worker details. The onsite worker details are updated during each cycle of the data fetch process.

(1)btnsubmit < - SubmitButton
(2)csv < - comma separated values
(3)ts < - timestamp
(4)fetch csv file every 5 seconds
(5)split the received data into rows
(6)create an array of rows
(7)remove the first two entries of array displaying ts and reader details (not required in table)
(8)join the array back into rows
(9)get each cell entry separately using Papa.parse() function
(10)if (EPC is repeated) then
(11)  remove duplicate entry
(12)  else
(13)  add entry to table
(14)end
(15)add a column for time using add DataToTable()
(16)if (EPC read) then
(17)  start function showtime()
(18)  show time in the format 00: 00: 00
(19)   else
(20)   Return to step 9
(21)End
(22)if (time is equal to shift ending time) then
(23)   click btnsubmit automatically (visibility: hidden)
(24)   data is submitted
(25)   reset table
(26)   data stored in firebase backend
(27)   else
(28)   keep receiving data
(29)End

The variables such as worker’s name, unique EPC product code, and timestamp show the present worker’s status at the construction site. Initially, the data were received from the Impinj software in the form of a CSV file which is fetched every 5 seconds to keep the database updated. As the tag is read in every time step, the multiple EPC entries are merged to remove the duplicate entry in reading. At the end of each shift, the workers' EPC and total time fetched will be used as their on-site working time to calculate their daily wages. Also, the supervisor will have an automatic update of the shift ending time of each worker. Finally, the updated data will be automatically submitted to the firebase database. Figure 21 shows the developed web page for workers’ personal details and the history of the authorized workers, which are stored in the firebase database record.

This web page will provide complete personal details such as the worker’s name, address, and contact details. The supervisor can also access the details of the worker by inputting his unique EPC code. Thus, this cloud-based firebase app provides the security and authenticity for the real-time monitoring of workers.

The current on-site details page is shown in Figure 22. The page will display different data fields such as the timestamp, EPC product code, workers’ location according to different reader antennas deployed in different working zones, and their calculated working hours. The timestamp shows the date and time when the worker enters the location. Other functionalities such as access control and alarm integration may be added further to restrict workers from entering dangerous zones.

4. Real-Time Testing of IoT-Enabled UHF-RFID System

For real-time testing and to validate application usage, the developed “IoT-Enabled UHF-RFID system” was tested at the construction site of B-block, Thapar Institute of Engineering & Technology, Patiala. The workers’ location and time information has been fetched and stored in the cloud-based firebase database through the designed middleware. For this, the designed tags were pasted over the helmet of the construction workers to monitor them, as shown in Figure 23. Furthermore, the complete measurement setup (zoomed) deployed at the real-time construction site for tracing the workers is shown in Figures 24(a)–24(c). The reader antenna was installed in the working zone to read the tagged workers’ status in the testing setup. However, in the practical setup, multiple antennas may be deployed to cover the entire working area or multiple working zones so that the tags do not get blocked by the reader signal and go unidentified.

Figure 25(a) shows the tracked record of the workers on the item test software, and Figure 25(b) shows the real-time on-site workers’ status displayed on the developed web application. The experiment was repeated three times, and it was observed that all of the workers were accurately tracked and monitored correctly. Furthermore, Figures 26(a)26(c) showed the on-site details of the remaining workers when few of them left the construction site. Thus, the experiment showed good accuracy for the monitoring and tracking of workers, showing the designed system’s robustness. Hence, the proposed tag antenna is a potential candidate to provide tracking solutions for the workers at complex job sites and, thus, for IoT applications.

5. Conclusion

In this work, a standalone RFID system is designed in which an RFID tag antenna equipped with WTMiddleware is developed to provide real-time tracking access control solutions for the workers at risky job sites. For this, a novel tag antenna design with good performance on metallic objects and ABS-based safety helmet is presented. Here, the microstrip patch-based folded dipole configuration is adopted for the tag antenna to exhibit conjugate impedance with respect to the Alien Higgs-3 RFID chip. Here, the information stored in the tag regarding workers in a real-world scenario was processed and accessed by a web-based application through developed WTMiddleware. Thus, RFID readers, tags, and middleware are combined to form a system for registering the workers and monitoring their on-site working status for providing digital worksite visibility of a real job site scenario. Hence, using the developed “IoT-enabled UHF-RFID system” offers a complete solution for the tracking of the workers and effective access control at the real construction job sites.

Data Availability

No new data were created or analysed in the study.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

The first Author, Aarti Bansal, is highly thankful to the Department of Science and Technology (DST), Government of India, for financially supporting the research work under the women scientist scheme (WOS-A) vide reference no. SR/WOS-A/ET-78/2018.