Development of microcontroller-based draft measuring system using Xbee technology

Abstract

Numerous wired data acquisition systems were developed in the past, for measurement of draft, which were often characterized by complexity and operational difficulty. Consequently, it is necessary to upgrade the exiting draft measurement technique of an agricultural implement. Therefore, this study was conducted to develop a microcontroller-based draft measurement system send the processed data to on-board stationary platform through wireless communication. This system includes (a) transmitting unit to transmit the measured draft value via Radio frequency based Xbee module, and (b) receiving and displaying unit to receive transmitted data from the transmitting unit and display it on LCD. Three load cells were attached to a three-point linkage dynamometer to sense the force exerted by the implements. The measured draft was wirelessly transmitted to the dashboard near to the tractor driver seat using Xbee technology. The draft measurement system was tested under actual field conditions. The results showed that the variations between the draft obtained from the developed draft measuring system and from the data logger were 6.5% and 7.16%, at 2.5 and 3.5 km/h, respectively. There was no statistically significant difference between the draft values obtained from the developed draft measuring system and the data logger. Observed data informed that the developed system was accurate, reliable and quick response in nature. The developed system can be implemented for the measurement of draft force for research and education purposes.

Keywords:

• Xbee technology
• microcontroller
• draft measurement system
• wireless system

Reviewing Editor:

• Pham D T, University of Birmingham, United Kingdom

Subjects:

• Agricultural Engineering
• Agriculture & Related Industries
• Agriculture

1. Introduction

A tractor is used as the prime power source in agricultural operations. For tillage, the drawbar power is one of the most important factors for agricultural implementation (Sprawka et al., 2023). Knowledge and information about drawbar power are essential for agricultural operations. This refers to the maximum pulling strength of a tractor at its hitching point for the implements. This knowledge is crucial for selecting compatible implements, monitoring field performance, improving efficiency and ensuring safety (Kathirvel et al., 2000; Usaborisut and Prasertkan, 2018). For effectively unitizing the power of tractor, it is important to the tractor power to be used most effectively, it is crucial to pair it with suitable implements. This matching ensures that the tractor’s output is fully utilized, maximizing its efficiency across various agricultural tasks (Sahu & Raheman, 2008). Drawbar measurements are important in tractive performance research. Properly aligning tractor implements assists in determining implement specifications such as depth and width; and aids in selecting the appropriate size of the tractor for specific tasks (Patel et al., 2012). Mehta et al. (2011) developed a decision support system to choose whether to match an implement to the tractor or to select a tractor with an implement. Numerous instrumentation systems have been developed to gauge the drawbar power of a tractor during implementation. However, the drawback is that most of these systems are customized for particular tractor models, limiting their adaptability across different machinery setups.

Three-point linkage dynamometers utilized for measuring tillage implement draft can be categorized into two primary groups (Chaplin et al., 1987). The first group comprises systems where transducers are positioned on a frame between the implement and the tractor, while the second group encompasses integral systems where the dynamometer arms are adapted to incorporate the transducers. Several researchers have developed instrumented mounted frames that are connected to three-point linkage dynamometers to measure the implement force. These frames employ transducer pins or strain gauges to measure the force within the link, thereby enabling precise measurements of the force exerted by the implement (Clark & Adsit, 1985; Godwin et al., 1993; Reece, 1961; Reid et al., 1985). Baker et al. (1981) developed a load cell-mounted frame to measure the forces exerted by the implements accurately. A dynamometer comprising a chassis, sensing unit and recording system has been designed and developed (Alimardani et al., 2008). They utilized strain gauges to detect the load applied by the implement, with data collection facilitated by a data logger. Many drawbar transducers have been developed to measure the drawbar power of an implementation (Kocher & Summers, 1987; Roeber et al., 2017). The draft measuring system was also developed using extended octagonal rings transducers and strain gauge-based hitch pins to measure draft force (Kirisci et al., 1994; Leonard, 1980; Tessier et al., 1992, Upadhyay et al., 2022). These systems then utilize commercial data loggers to record and analyze the acquired data. However, these methods have certain limitations. This necessitates the use of expensive commercial data loggers to record output data, which can be cumbersome. Furthermore, these devices lack digital data display features and add complexity to the data management.

In several studies, mathematical models were additionally formulated to forecast the draft force (Arvidsson et al., 2004; Grisso et al., 1996; Taniguchi et al., 1999). The draft magnitude is influenced by factors such as soil type and its condition, tool characteristics, working speed and depth (Gill & Vanden Berg, 1967; Kydd et al., 1985). Several researchers have also developed empirical polynomial and linear regression models. These models aim to predict the draft of tillage implements based on empirical data and analyses (Grisso et al., 1996; Godwin et al., 2007; Kheiralla et al., 2004; Mahore et al., 2024). Nevertheless, many of these models encounter issues related to multicollinearity that can affect their accuracy. Additionally, their applicability tends to be confined to specific soil types and implementation conditions for which they were originally developed.

Options for commercial devices that can measure and display draft forces are quite limited and tend to be expensive. Some wire-based data acquisition systems exist for measuring the draft of a tractor, yet they often prove to be complex to operate and can lead to operational issues during use. Therefore, it is necessary to develop a microcontroller-based draft measuring system that wirelessly sends it to the dashboard of the tractor driver.

2. Materials and methods

A microcontroller-based draft measurement system was developed using the Xbee wireless technology. This system was specifically designed to measure the draft of mounted-type implements using a dynamometer. Kumar et al. (2016) designed and developed a three-point linkage dynamometer specifically for measuring drafts at Indian Institute of Technology (IIT), Kharagpur. The developed system integrated three load cells to accurately sense the load exerted by the implement during operation. The load cell detects force and transmits it to the Arduino unit of the transmitter unit to give draft as an output. The Xbee unit of the transmitter unit sends this draft to the Xbee unit of the receiving unit to display. This draft measuring system consists of an Xbee unit, Arduino Uno, and an amplifier.

2.1. Three-point linkage dynamometer

A three-point linkage dynamometer (Figure 1) was designed and developed to measure the drafts of different types of mounted implements at IIT, Kharagpur. The dynamometer was universal in nature, and therefore, it may adapt to fit any tractor-implement combination. The dynamometer features a detachable sensing unit that fits the load cell. The front end of this frame was connected to the tractor, whereas the rear end was affixed to the implementation. Three load cells were employed to sense and measure the draft of the implementation. These cells were positioned within three sensing units, all of which were affixed to a dynamometer for measurement.

Figure 1. Three-point linkage dynamometer: (1) sensing unit; (2) arms; (3) body; (4) load cell.

2.2. Development of the microcontroller-based draft measuring system

Wire-based data acquisition systems for measuring tractor performance can be complex and problematic during operation. Therefore, there is a growing need for wireless data-acquisition systems. These wireless systems offer easier management and operation, eliminating the challenges posed by the wired setups. A microcontroller-based draft measuring system was developed to measure the draft and display near to tractor driver seat. Wireless communication was performed by applying the Xbee technology. The embedded system was composed of two primary units: a draft measuring and transmitting unit, for collecting and transmitting draft data and a receiving and displaying unit the transmitted data. The units consist of microcontroller (Arduino Uno), Xbee shield, Xbee module for wireless communication, LCD for display and amplifier. Configuring the Xbee modules in both units establishes a wireless network between the modules, allowing smooth communication between the draft measuring and transmitting unit, and the receiving and displaying unit.

2.2.1. Configuration of Xbee module

The Xbee module of the developed draft measuring and transmitting unit system provides secure access to networked radio over the air. To wirelessly transmit data, the Xbee module needs to be configured for the transmission control protocol (Boonsawat et al., 2010; Desnanjaya et al., 2020; Kabir et al., 2014; Kumar, 2016). Both the Xbee modules were configured using Digi’s X-CTU software. Prior to the configuration, the Xbee modules were placed on an Xbee shield, and both modules were connected to a laptop using a USB cable (see Figure 2). One module function as a router (receiver), and the other module operates as a coordinator (transmitter). The addresses of both Xbee modules (Table 1) were set to be identical. Numerous other settings were adjusted within the XCTU software according to specific needs. The configuration window for the XBee module is illustrated in (Figure 3).

Figure 2. Configuration of Xbee module.

Figure 3. Configuration window Xbee model: (a) Transmitter; (b) Receiver.

Table 1. Technical specification of sensor used to measure different parameters.

S. No	Sensor	Manufacturer	Model	Features	Application
1.	Load cell	Monad Electronics (India)	MN-TC55	1. Full strain gauge bridge 2. Capacity: 20 kN 3. Rated output: 2.0 mV/V	To measure the draft
2.	Hall‐effect sensor	Sun Robotics (India)	A3144	1. Type: Magnetic Hall sensor 2. Output current: 25 mA 3. Polarity: Unipolar	To measure forward speed
3.	Xbee module	Digi (USA)	S2C	1. Data rate: RF 250 Kbps 2. Range: Indore-60 m and outdoor-1200 m	To transmit draft, transmitting unit to receiving unit

2.2.2. Development of draft measuring and transmitting unit

A wired data acquisition system poses challenges in handling and collecting information from the sensors. Therefore, it is necessary to create a transmitting unit to wirelessly transfer data from the sensor. The transmitting unit was designed using the Xbee module-based technology. It comprises a microcontroller (Arduino Uno), Xbee module, Xbee shield, amplifier (INA125) and load cell pin. The amplification process involved connecting three amplifiers, each corresponding to a load cell, to amplify their respective signals. Three load cells were linked to the load cell pin to detect and measure the applied load on the implement during the field operations. The experienced load was amplified by the amplifier to minimize errors, and then transmitted to the microcontroller. The microcontroller reads and processes the signal received from the amplifier and converts it into a digital value before transmitting it to the Xbee module of the receiving unit. The Xbee module connected directly to the microcontroller using jumper wires. The data received from the microcontroller forwarded to the Xbee module of the receiver unit. The developed transmitting unit and circuit diagram of this unit are shown in Figures 4 and 5, respectively.

Figure 4. Developed draft measuring and transmitting unit: (1) microcontroller; (2) Xbee module; (3) amplifier; (4) load cell pin.

Figure 5. Circuit diagram of draft measuring and transmitting unit.

2.2.3. Development of receiving and displaying unit

A receiving and displaying unit was to receive and display the signals sent to the transmitting unit. It includes a microcontroller, Xbee module, Xbee shield and LCD. The Xbee module receives the signal from the transmitting unit and sends it to the microcontroller. The microcontroller reads and processes the signal from the Xbee module to be displayed on the LCD. The developed receiving and displaying units and circuit diagram of this unit, depicted in Figures 6 and 7, respectively. Flowchart of the developed wireless embedded system is shown in Figure 8.

Figure 6. Receiving cum displaying unit.

Figure 7. Circuit diagram of receiving and displaying unit.

Figure 8. Flow chart of developed PTO torque transducer.

2.3. Calibration setup for load cell

Three load cells (Table 1) were required to measure the draft force of implementation. Out of which, two load cells were positioned in the lower link and the remaining one in the upper link. The upper link and lower link’s load cells experienced tension and compression force, respectively. Prior to field, it was essential to calibrate both load cells for compression and tension forces. A microcontroller-based electronic circuit developed to calibrate the load cell. The calibration of the load cell was aimed at assessing fluctuations in sensitivity and linearity across a range of different dead loads.

2.3.1. Microcontroller-based electronic circuit

The electronic circuit consisted of a microcontroller (2 KB SRAM, Smart Projects, Italy) and an amplifier (INA125, 2.7 to 36 V, Texas Instruments, India). INA 125 was linked to the microcontroller using a jumper wire, whereas its other pins were connected to the load cell integrated into the calibration setup. An amplifier employed to enhance the signal produced by the application of compression and tension loads. The amplified signal from the amplifier was transmitted to the microcontroller for processing and was converted into a readable format. The microcontroller-based electronic circuit used for the calibration is shown in Figure 9.

Figure 9. Developed calibration circuit.

2.3.2. Calibration setup for tension force

The calibration setup consisted of a load crane, digital load cell, calibration circuit, laptop and load cell. The two load cells were calibrated by connecting them in the calibration setup. The load cell was fastened at one end to the upper section of the load crane, whereas its opposite end was linked to the lower part of the crane. A digital hanging type weighing scale was positioned between the load cell and load crane to display the applied load on the load cell. A load crane was employed to apply tension force to the load. The output signal from the load cell was observed in an analog form. The analog signal is directed to the amplifier for signal amplification before being transmitted to the microcontroller, where it is converted into a digital value. Alterations in the digital value (sensor reading) corresponding to the applied tension load were recorded on a computer. Graphs (Figure 10(a,b)) were then generated, depicting the relationship between the applied load and sensor values. The observations revealed a linear relationship between the sensor value and applied load. The calibration setup that demonstrates this relationship is illustrated in Figure 11.

Figure 10. (a) Digital output versus load (tension force) for load cell 1(RC01C). (b) Digital output versus load (tension force) for load cell 2 (RC02C).

Figure 11. Calibration setup for tension load.

2.3.3. Calibration setup for compression force

Another calibration setup (Figure 12) was developed for the compression force. This setup comprised a hydraulic jack, load cell, weighing platform, load cell and laptop. A hydraulic jack was positioned on a weighing platform, and the force exerted on the platform was displayed using a digital weighing balance. The load cell was situated between the hydraulic jack and the iron bar within the setup. A hydraulic jack was used to apply a compressive load to the load cell. The output of the load cell displayed on a laptop screen. The calibration curve for the compressive force is shown in Figure 13.

Figure 12. Calibration setup for compression load.

Figure 13. Digital output versus load (compression force) for load cell 3 (RC03C).

2.4. Measurement of tractor forward speed

A Hall sensor (Table 1) was used to measure the forward speed of the tractor during field operations (Shrivastava et al., 2023; Chouriya et al., 2023; Kushwah et al., 2024; Gupta et al., 2019; Nataraj et al., 2021). The Hall sensor is positioned on the rear wheel of the tractor. To measure the RPM accurately, small magnets were uniformly affixed (Figure 14) to the wheel rim. As the wheel rotates, the Hall sensor detects these magnets and generates voltage signals in response to the magnetic field. These sensor outputs were then fed into a microcontroller (ATMEGA328; Smart Projects, India). Using an Arduino program developed specifically for speed measurement, the system calculated the speed of the tractor based on these signals.

Figure 14. Forward speed measurement setup.

3. Testing of developed microcontroller-based embedded system

Testing of the developed microcontroller-based embedded system was conducted in the research field of the Department of Agricultural and Food Engineering, IIT Kharagpur. A 2 WD tractor (Samrat 4410, 44 hp, TAFE, India) was used to test the developed draft measuring system. The code was scripted using the Arduino IDE software and uploaded to the memory of both the draft measuring cum-transmitting unit and the receiving cum-displaying unit to initiate their functions. The draft measuring cum-transmitting unit was affixed to a three-point linkage dynamometer using adhesive tape. During the operation, the load cell senses the load, which then transmitted to the amplifier within the transmitting unit. The amplifier’s role was to enhance the accuracy and quality of the data received from the load cell, thus minimizing potential errors in the transmitted information. Following amplification, the strengthened signal directed to the microcontroller for reading, processing and conversion into digital values. This digital signal is then relayed to the Xbee module, facilitating its transfer to the Xbee module of the receiving unit. The Xbee module within the receiving unit transmitted the received signal to the microcontroller. Subsequently, the microcontroller processes this signal to facilitate its display on the LCD screen. Figure 15 shows the field test conducted on the developed draft measuring system using a cultivator. A data logger was used to measure the draft of the implementation to validate the observed draft from the developed embedded system.

Figure 15. Field testing of developed embedded system.

An ANOVA test was also conducted to determine how the forward speed affects the draft of the implement. One-way ANOVA results demonstrated a significant effect of the forward speed of the tractor on draft value. The draft value obtained from the developed microcontroller-based draft measuring system was compared to the observed draft value recorded by the data logger. The deviation between the draft values was depicted using the mean absolute deviation (MAD) and mean absolute percentage error (MAPE). The MAD indicates the average variation or average difference between individual values and their means. Mean absolute deviation error (MAPE) measures the accuracy of a system by illustrating the average magnitude of errors as a percentage of the actual values (Khair et al., 2017; Owolabi et al., 2021; Fitriyah & Budi, 2019; Hensh et al., 2021). (1) $$\mathrm{\text{MAD}}=\frac{1}{n}\mathrm{ }\sum _{i=1}^n\left|\frac{T_e-T_d}{T_e}\right|$$(1) (2) $$\mathrm{\text{MAPE}}=\frac{1}{n}\mathrm{ }\sum _{i=1}^n\left|\right|T_e-T_d|$$(2)

4. Results and discussion

4.1. Effect of forward speed on draft

An experiment was conducted to assess the developed microcontroller-based draft measuring system under field conditions. The evaluation involved the utilization of nine tine cultivators at speeds of 2.5 and 3.5 km/h. Figure 16 illustrates the fluctuations in the draft with respect to the forward speed of the tractor. The graph demonstrates that as the forward speed of the tractor increased, there was a corresponding increase in the draft of the implement. The increase in the draft of the implement was attributed to an increase in the acceleration of soil particles. Statistical analysis was performed to assess the influence of forward speed on the draft of the implement. The ANOVA results indicated a significant effect of the forward speed on the draft of the implement (p < 0.05). The maximum draft values recorded were 301.35 kg at 2.5 km/h and 335.12 kg at 3.5 km/h, respectively.

Figure 16. Variation of observed draft at speeds of 2.5 and 3.5 km/h for nine tine cultivators.

4.2. Comparison of draft measured by embedded system and data logger

The draft values observed from the developed embedded system were compared with those acquired from a data logger (HBM, USA). It was noted that at speeds of 2.5 and 3.5 km/h, the average torque values registered were 321.49 and 359.18 kg, respectively, for the data logger. At speeds of 2.5 and 3.5 km/h, the observed draft values of the developed torque transducer deviated by 6.5% and 7.16%, respectively, from the draft values of the data logger, as depicted in Figure 17.

Figure 17. Comparison of draft measured by embedded system and Data logger at 2.5 and 3.5 km/h.

5. Statistical analysis’s outcomes

A statistical analysis was performed to evaluate the performance of the developed draft-measuring system. At speeds of 2.5 and 3.5 km/h, the MAPE stood at 3.6% and 2.0% correspondingly, while the MAD values were 4.45 and 1.21 kg, respectively. Robinette and Daanen (2006) proposed that a developed system would be considered acceptable if the MAD value was less than 10. Many scientists (Lee et al., 2014; Cortez & Morais, 2007; Habibnia et al., 2019) have supported the value of the MAD as a reliable measure to determine the acceptability of a system’s performance. The developed embedded system consistently displayed high application accuracies, fluctuating between 96.4% and 98%, averaging at a commendable 97.2%. One-way ANOVA was performed to evaluate the influence of the forward speed of the tractor on the draft. The findings indicate a notable and significant effect of the forward speed of the tractor on the draft. Naderloo et al. (2009) and Salahloo et al. (2021) also demonstrated that forward speed notably influences the draft of the implement.

6. Conclusions

An embedded system utilizing Xbee technology has been specifically engineered to measure drafts and wirelessly transmits these data under dynamic conditions. The developed system was validated under field conditions using a data logger, and the results proved to be satisfactory. The performance parameters suggest that the developed microcontroller-based embedded system can measure drafts effectively and accurately. The velocity at which the tractor moved significantly affected the draft force experienced by the implementation.

The findings indicated that there were differences of 6.5% and 7.16% between the draft values measured by the developed system and those recorded by the data logger at speeds of 2.5 and 3.5 km/h, respectively. Additionally, the MAPEs for the observed draft values from the developed system compared to the data logger were 3.69% and 2.0% at 2.5 and 3.5 km/h speeds, respectively. The developed system demonstrated the capability of measuring draft values with an accuracy of approximately 93%. Statistical analysis indicated that there was no significant difference between the draft values obtained from both systems. This system has a transmission range of up to 100 m through the air. It offers convenience, simplicity, reliability, and accuracy in operation, making it suitable for use in various agricultural tractors.

Author contributions

Arjun Chouriya: Conceptualization, Methodology, Formal analysis, Writing-original draft Software.

V. K. Tewari: Funding, Supervision.

Peeyush Soni: Writing-review and editing.

Naseeb Singh: Data curation, Visualization.

Pradeep Kumar: Writing-original draft.

Disclosure statement

The authors declare that they have no known competing financial interests or personal relationships that could have influenced the work reported in this study.

Data availability statement

Data will be made available on request.

Additional information

Funding

This research was funded by the Agriculture and Food Engineering Department, IIT Kharagpur, India (Institute ID IR-O-U-0573).

Notes on contributors

Arjun Chouriya

Arjun Chouriya Research Scholar, IIT Kharagpur.

V. K. Tewari

V. K. Tewari Director, IIT Kharagpur.

Peeyush Soni

Peeyush Soni Prof. IIT Kharagpur.

Naseeb Singh

Naseeb Singh Scientist, ICAR Research Complex for NE Hill Region, Umiam, India.

Pradeep Kumar

Pradeep Kumar Research Scholar, IIT Kharagpur.