Introduction

Concrete is widely employed in construction industry as a durable, affordable, and structurally functional material. [1,2,3]. Despite its widespread use, estimation of the time needed to obtain specific strength is challenging as it is influenced by various factors. Construction operations, such as removal of shoring systems, slipforming operations, and stressing of post-tensioned tendons, could potentially be optimised with an accurate assessment of the early-age concrete strength [2, 4,5,6,7,8,9,10,11,12,13,14,15,16,17]. Monitoring the early-age concrete strength could be also used to determine the 28-day compressive strength and ensure concrete reaches its expected design strength.

To assess the mechanical characteristics of concrete and make choices on on-site construction, concrete cylinders are often cast and cured on site. However, the cylinders do not realistically reflect the properties of the in-place concrete because the curing regime and boundary circumstances of concrete samples are drastically different from the in situ curing conditions [5, 18]. In addition, low temperatures may result in insufficient strength development and therefore longer curing times to reach the necessary concrete strength [1]. It is essential to correctly anticipate the concrete strength during the early phases of construction to avoid contractors acting prematurely as inadequate strength can cause problems such as cracking, performance, and durability issues [4, 8]. Thus, real-time monitoring of the short-term strength development is necessary for the quality and safety assurance of concrete buildings. Several structural collapses because of early stripping of formwork have been reported, such as the collapse of the Skyline Tower in Fairfax County in 1973 [19] and a cooling tower in Willow Island in 1978 [20, 21].

The maturity method, which is a non-destructive method, can be employed to predict the early-age strength development (less than 14 days) of concrete based upon the temperature history of the concrete. Maturity curves are developed for various concrete mixes based on the early-age temperature rises in the hydration process of the same mix. This approach is predicated on the relationship between compressive strength and rate of hydration, which is a function of internal temperature rise. Monitoring this shift will thus offer important information about the hardening process which can be used to evaluate the short-term strength.

In investigations conducted on the maturity method to date, researchers have proposed various maturity functions [22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39], some of which have been documented in numerous standards and guidelines [40,41,42,43,44,45]. Devices for measuring maturity, such as thermocouples and wired temperature and maturity loggers, as well as Internet of Things (IoT)-based systems (cloud computing platforms linked to wireless sensors), may be used to monitor the strength of concrete [4, 18, 46, 47]. The drawbacks of conventional maturity meters include manual operation onsite and the requirement of a software for data extraction. The limitations are corrected and enhanced in the IoT-based systems to automate the temperature observations and maturity index computations. Cloud-based systems prevent data loss and facilitate live monitoring from remote places [18].

Recently, CDS has developed an IoT-based system for real-time monitoring of early-age concrete strength using the maturity method. The system has several new features including sensor network signal range greater than 500 m and longer sensor battery life. Furthermore, the sensors can provide both short- and long-term monitoring of other parameters, such as relative humidity, movement, and vibration, aiding in minimising the maintenance expenses and extending the lifetime of structures. Nevertheless, CDS collaborated with Bond University to only investigate the feasibility of their sensor system for real-time monitoring of the early-age compressive strength of concrete. The main objective of this paper is to present the details and performance of the system with respect to the early-age concrete strength monitoring. Investigating the other functions of the sensor system are beyond the scope of this paper.

Monitoring systems

Thermocouples, wired temperature and maturity loggers, wired concrete sensors with external wireless transmitter, and fully embedded wireless concrete sensors are the available maturity measurement systems [48, 49]. Figures 1 and 2 present an illustration and typical examples of each system, respectively. Although use of thermocouples is very labour-intensive, they are commonly utilised in the concrete industry since they are inexpensive and provide a suitable temperature measurement range. Maturix [50] and HOBO [51] are some of the commercially produced items in the market. Wired temperature and maturity loggers have a connecting wire that extends to a hand-held device to download the measured data. These systems require physical transfer to a computer to perform data analysis. Command Center [52], intelliRock [53], and Con-Cure Nex [54] are some of the commercially produced items in the market. The necessity for on-site data collecting at construction sites was addressed by the design and development of wireless data transmitters. The transmission can be conducted: (1) using a computer with an internet connection; (2) using a local wireless network to transmit data to a cloud, or (3) utilising networks like LTE or Sigfox to send data directly to a cloud. Maturix [50], Command Center [52], intelliRock [53], Concremote [55], Converge Signal [56], Extract Technology [57], HardTrack [58], and Lumicon [59] are some examples of commercialised systems available in the market.

Fig. 1
figure 1

Illustration of monitoring system (adapted from [49]). a Concrete thermocouple; b wired temperature and maturity loggers; c wired concrete sensors with external wireless transmitter; d fully embedded wireless concrete sensors

Full size image
Fig. 2
figure 2

Example of commercially available monitoring system. a Maturix thermocouple [50]; b intelliRock wired concrete temperature and maturity loggers [53]; c lumicon wired concrete sensors with external wireless transmitter [59]; d smartRock fully embedded wireless concrete sensors [71]

Full size image

Many studies have employed IoT-based systems in various applications in construction industry [60,61,62,63,64]. Based on established maturity functions, limited research has constructed Internet of Things (IoT)-enabled systems to examine the early-age concrete strength monitoring [18, 65,66,67,68]. John et al. [65] developed an IoT-enabled system to monitor early-age strength of cementitious materials. Yikici et al. [66] developed a LoRaWAN wireless IoT-based system to monitor and calculate concrete maturity. An IoT-based system was developed by Lim et al. [68] to record the internal temperature in concrete members using DS18B20 temperature sensors and microcontroller Raspberry Pi B3 +  Data were transmitted using Apache’s NiFi and MiniFi. A SQL database server was used to store data. A system consisting of Wi-Fi microcontrollers and temperature sensors was developed by Olaniyi et al. [67]. This system was linked to a web-based application through DS18B20 sensors to record the temperature and transmit the data to the ESP32 Wi-Fi microcontroller. More recently, John et al. [18] introduced a similar cloud-based system to monitor the early-stage strength gain which consisted of DS18B20 temperature sensors an ESP8266 Wi-Fi microcontroller. Even if project managers are not on the construction site, automated data analysis on the cloud and alerts are used to notify them when necessary. However, there is still a chance of on-site damage to the stretched wire out of concrete parts. Moreover, the wireless system setup and configuration processes might be challenging. Application of these systems also depends on cellular network connectivity.

Nowadays, fully embedded wireless concrete sensors and loggers have become readily accessible. [69]. The recorded temperature data are saved to an internal memory placed within the sensors. The data are transmitted using wireless communication systems such as LoRa [70], Zigbee, Bluetooth Low Energy (LE), or Wifi [48, 69]. A tablet or smartphone, wirelessly connected to the sensor, can be used to collect data and perform data analysis through available mobile applications. Alternatively, use of a local gateway eliminates the need to be on site and can be used to securely transmit data to the cloud platform. For other wireless communication protocols, an external gateway is required as these systems are not supported by mobile devices. Converge Signal [56], HardTrack [58], Concrete Sensors [71], SmartRock [72], and vOrb [73] are commercialised products available in the market.

Research significance

The maturity method can be used to optimise numerous applications in concrete structures. At the construction site, the removal times and method of striking formwork are chosen at random due to absence of a widely accepted approach. Accurately determining the speed of strength development in concrete is essential to find out the shortest demoulding time [74] and thereby achieve significant cost reductions [75,76,77,78]. This is more crucial when mass concrete in large-scale structural elements is used in construction. Precast components are another application that may benefit from the maturity process. Time delays in the production of these components, which have an impact on their cost-effectiveness, are the primary concern of precast manufacturing. Appropriate estimation of the short- and long-term mechanical characteristics of the precast components is necessary to reduce production time [79]. Technically, the main problem with post-tensioned (PT) construction is the absence of standardised procedures to evaluate the in-place strength of concrete used in PT concrete slabs [5]. The procedure of initial and final stressing in PT concrete slabs requires early-stage strength determination [80]. An exact assessment of the in-place strength determines the appropriate age for tensioning the tendons. To ensure long-term safety, monitoring of early-age strength could also be used as an indicator to detect factors affecting later age strength gain of concrete. Issues in mix design, quality of materials used in concrete, curing conditions, etc. could be identified as early as possible and remedial action performed as required. To address the above-mentioned issues, conventional maturity meters are not appropriate for the in situ monitoring of strength. As a result, the automation of the entire maturity process is quite necessary for real-time monitoring of in situ concrete [18, 81]. This research is one of the few papers available in the literature that uses a novel and cost-effective IoT-based system to automate real-time monitoring of in-place concrete and accurately predict the early-age strength [18, 82, 83].

Details of the system

A prototype of the system was supplied by CDS to evaluate its performance at Bond University. Details of the proposed system are described in this section.

Concrete maturity method

The system is based on the reliable and widely used maturity function given by Freiesleben Hansen and Pedersen [28] to predict the early-age compressive strength of concrete. The equivalent age expression by Freiesleben Hansen and Pedersen is one of the most used functions which are based on the Arrhenius equation. The expression is standardised in most standards such as ASTM C1074 and the European code. As it uses apparent activation energies, the Arrhenius function produces more reliable estimates of the early-age compressive strength. A thorough explanation of activation energies can be found in [84, 85], and their determination procedure is available in the ASTM standard practice. The equivalent age in the maturity function is given by:

$$t_{{\text{e}}} {\mkern 1mu} = {\mkern 1mu} \sum\limits_{0}^{t} {e^{{ - \frac{{E_{a} }}{R}\left( {\frac{1}{{T\, + \,273}} - \frac{1}{{T_{{\text{r}}} + 273}}} \right)}} } \, \times \,\Delta t$$
(1)

where te is the equivalent age at a specified temperature T (days or hours); E is the apparent activation energy (J/mol); R is the universal gas constant (8.314 J/mol K); T is the average temperature of the concrete during interval Δt (℃); and Tr is reference temperature (℃). The value of activation energy can be obtained as follows:

$${\text{for}}\;T\, \ge \,20\,^\circ {\text{C}}{\mkern 1mu} {\mkern 1mu} {\mkern 1mu} {\mkern 1mu} E\, = \,\text{33,500}\;{\text{J}}/{\text{mol}}$$
(1a)
$${\text{for}}\;T\, < \,20\,^\circ {\text{C}}\,\,\,\,~E\, = \,\text{33,500}\, + \,1470\, \times \,\left( {20\, - \,T} \right)\,{\text{J}}/{\text{mol}}~~$$
(1b)

Tr is the standard curing temperature which is considered as 20 °C in many regions. In Australia, it is 23 °C and 27 °C in temperate zones and tropical zones, respectively [5].

Once the equivalent age has been established, the strength at a certain age can be obtained using the three-parameter equation (TPE) [30]:

$$S = S_{\infty } e^{{ - \left( {\frac{\tau }{{t_{e} }}} \right)^{\alpha } }}$$
(2)

where S is the strength at age te (MPa); S∞ is the limiting strength (MPa); τ is the characteristic time constant (hours); and α is the shape parameter. The parameters need to be calibrated using concrete cylinder tests for different concrete mixes.

The developed system

As noted earlier, the developed system is an IoT-enabled system. IoT is a network of interlinked sensor nodes that are given unique identifiers. Sensor-recorded data can be transferred over the network through a wireless link to a cloud-based platform [86]. Due to the sensor nodes’ continued connectivity with one another and the Internet, the IoT system can be efficiently utilised to automate the real-time monitoring of different factors including concrete internal temperature to estimate the early-stage strength gain.

The framework of the IoT-enabled system is shown schematically in Fig. 3, and the system, consisting of the gateway, sensor, and LoRa Antenna, is presented in Fig. 4. In the proposed system, DS18B20 waterproof temperature sensor is integrated into the sensor to measure temperature under wet conditions. Table 1 gives specifications of the sensor. The sensor can measure temperatures ranging from − 20 to 70 °C with the accuracy being ± 0.8 °C, which satisfies the requirement of ASTM C1074 [41] to be within ± 1 °C. Figure 5a presents a photograph of the sensor without its protective case. The gateway is designed and manufactured by CDS and is a long-range (LoRa)-based communication set with fully integrated technology to allow the onboarding of different sets (temperature, humidity, movement, and vibration). The gateway does have internal rechargeable batteries; however, the system is designed to be powered continuously from mains power. Each sensor must be paired with the gateway using a magnet. Specifically, by pressing the [CLICK TO PAIR] on the gateway’s touch screen (see Fig. 5b), the signal bars will start to flash until a device is found or pairing has time out. To power on the sensor, the magnet is subsequently placed to the top of the sensor for about 1 s (see Fig. 5c). Once the sensor has successfully paired to the gateway, data start to be displayed on the gateway screen (see Fig. 5d). The gateway communicates with the sensors via LoRa and is fully patent protected in its application to transmit data through concrete. The gateway sends data packets via Global System for Mobile Communications (GSM) to the cloud-based portal, allowing users to access from any smart device to view, download, and save a report of all concrete results. To predict the early-age strength, it is essential to develop and implement the maturity relationships into the portal. The maturity relationships of the specified concrete mix are calibrated with the actual strength data to ensure accurate real-time predictions. Based on analysed results, the portal also permits informed construction sequence decisions to be made, such as: when to limit back-propping, early formwork removal, transfer of prestressing to a structure, and efficient use of actual structural capacities to calculate the construction loads. The flowchart for implementing the monitoring system is presented in Fig. 6. The battery life of sensors is circa 7 years. However, sensors can be designed for site specific solutions to last shorter or longer, which also translates into smaller and lighter sensors, or larger and heavier ones.

Fig. 3
figure 3

IoT framework of the developed system

Full size image
Fig. 4
figure 4

Monitoring system

Full size image
Table 1 Specifications of DS18B20 integrated in the sensor
Full size table
Fig. 5
figure 5

a Details of the sensor; b gateway home screen; c pairing a sensor; d activity description on the gateway screen

Full size image
Fig. 6
figure 6

Flowchart for implementing the system

Full size image

Reliability of the sensors

The reliability of the sensor was also assessed by comparing the recorded temperature data with those recorded by an EL-USB-TC-LCD thermocouple data logger as presented in Fig. 7a. Detailed specifications of the logger can be found in EL–USB–TC–LCD Datasheet [87]. The data were downloaded by plugging the data logger into a PC’s USB port and using the EasyLog software. Figure 7b provides a comparison of the temperature recorded by the sensor and thermocouple data logger. As indicated, the recorded temperature data match well confirming the accuracy of the sensor.

Fig. 7
figure 7

a Thermocouple data logger; b comparison of the temperature recorded by the developed sensor and thermocouple data logger

Full size image

Prediction of early-age compressive strength

To evaluate the performance of the proposed system, three typical concrete mixes referred to as Mix 1, Mix 2, and Mix 3 were used to predict the early-age compressive strength. Based on the recorded internal temperature–time history data in the concrete by the sensors, the strength–age relationships of all concrete mixes were derived. The early-age compressive strength of concrete was then predicted using the derived strength–age relationships. Ultimately, the reliability of the system was evaluated by comparing the predicted compressive strength with the measured compressive strength of concrete at different stages.

Material properties

General-purpose (GP) cement, sand, and 10/20 mm coarse aggregates were used in this study to produce the concrete. Nominal maximum size of aggregate in the concrete was in accordance with AS1012.8.1 [88]. Water-to-cement ratios of the three concrete mixes ranged from 0.4 to 0.53. Table 2 gives the details of the three concrete mixes.

Table 2 Concrete mix design
Full size table

Strength–maturity relationship

Seventeen standard-sized cylinders (200 mm high and 100 mm in diameter) of each mix (total of 51 samples) were cast as per AS1012.8.1 [88]. To record the internal temperature of concrete samples at every 60 min, the sensors were embedded in two cylinders of each mix during casting. Figure 8a indicates an embedded sensor in a concrete sample, and Fig. 8b shows the casting concrete cylinders. Typically, field workers cast cylinders, which are then taken to a lab for testing. Due to varying curing circumstances, the qualities of the concrete in the cylinder and the concrete poured on-site may be different. The Australian Standard on specification and supply of concrete AS1379 [89] recommends concrete cylinder testing according to AS1012.1, AS1012.8.1, and AS1012.9 [88, 90, 91]. The standard method AS1012.8.1 [88] specifies a curing situation that is quite different from the curing circumstances of in situ concrete. The concrete cylinders are exposed to typical moisture conditions at constant temperatures, as specified by AS1012.8.1 [88]. However, according to the Australian Concrete Structures Standard AS3600 [92], in situ concrete and concrete cylinder samples shall be cured under identical circumstances. The phrase “identical circumstances” still needs an in-depth elaboration. According to De Carufel et al. [4], the field environment may not be indicative of the calibration (strength–maturity relationship) in a humid setting; hence, methods such as employing burlap and plastic covering during curing can be used. In this research, all concrete cylinders were cured using wet hessian bags (see Fig. 8c). The compressive strength of concrete cylinders was tested using an automatic compression machine for all three mixes, as shown in Fig. 8d, after 1, 3, 7, 14, and 28 days of curing (mean of three samples). All tests were undertaken following the guidelines prescribed by AS1012.9 [91]. The results of cylinder strengths for the three concrete mixes are presented in Fig. 9.

Fig. 8
figure 8

a Sensor embedded in the fresh concrete; b casting concrete cylinders; c cylinders cured under wet hessian bags; d compression test of a cylinder

Full size image
Fig. 9
figure 9

Compression test results for the three concrete mixes at various ages

Full size image

The temperature data received by the gateway from the wireless sensors are transmitted to the cloud network through LoRa wireless technology. The cloud network can be remotely accessed by multiple devices, such as laptops, tablets, and smartphones. The system automatically calculates equivalent ages for each mix based on Eq. (1) using average temperature. Figure 10 represents the equivalent age profiles of the three concrete mixes.

Fig. 10
figure 10

Equivalent age profiles for the three concrete mixes

Full size image

The strength–age relationships of each mix were derived by calibrating the parameters S∞, τ, and α. The compressive strength and corresponding equivalent age data at ages 1, 3, 7, 14 and 28 days for each concrete mix were used to conduct a regression analysis to obtain these parameters, as given in Table 3. The derived strength–age relationships can subsequently be employed for the early-age strength estimation.

Table 3 Parameters used in strength–age relationship
Full size table
Table 4 Actual compressive strength and predicted values by the system
Full size table

Strength comparison

For each of the three concrete mixes cast in “Material properties” Section, a set of fourteen concrete cylinders was further cast to validate the performance of the system. Similar to specimens used to develop maturity relationship, the sensors were embedded in two out of fourteen cylinders. Compressive strength tests at ages 1, 3, 7, and 14 days of curing were carried out for the remaining twelve cylinders. The mean compressive strength of three cylinders obtained from compression tests was compared with those predicted by the system. A typical display of the predicted early-age compressive strength on the portal for Mix 2 is shown in Fig. 11. The comparisons of the predicted values and actual test results for the three concrete mixes are presented in Figs. 12,13 and 14, and Table 4. The ratios of predicted to actual strengths varied from 0.93 to 1.08, for Mix 1 and Mix 3, respectively. The results demonstrate that the system satisfactorily estimated the early-age compressive strength of the investigated specimens.

Fig. 11
figure 11

Real-time monitoring of the early-age concrete strength displayed on the portal for Mix 2

Full size image
Fig. 12
figure 12

a Comparison of the predicted and actual compressive strength for Mix 1; b ratio of predicted/actual strengths

Full size image
Fig. 13
figure 13

a Comparison of predicted and actual compressive strength for Mix 2; b ratio of predicted/actual strengths

Full size image
Fig. 14
figure 14

a Comparison of predicted and actual compressive strength for Mix 3; b ratio of predicted/actual strengths

Full size image

Conclusion

This study proposes an IoT-based system to automate real-time monitoring of the short-term strength development of concrete. The developed monitoring system consists of a gateway, sensor, and LoRa Antenna. The performance of the system was investigated by calibrating strength–age relationships for three concrete mixes in the laboratory. The maturity relationships were further employed and implemented into the system to estimate the early-stage compressive strength of concrete. The system was found to satisfactorily predict the compressive strength of the mixes. The system sends data packets to the cloud-based portal which can be remotely accessed. The received data are analysed on the portal to facilitate making informed construction sequence decisions, and therefore, the system can be effectively used to optimise removal of formwork, estimate stressing stage of post-tensioned tendons, and shorten manufacturing time of precast elements.