Comparison between Different Types of Sensors Used in the ... - MDPI

sensors Article

Comparison between Different Types of Sensors Used in the Real Operational Environment Based on Optical Scanning System Wendy Flores-Fuentes 1 ID , Jesús Elías Miranda-Vega 2, *, Moisés Rivas-López 2 Oleg Sergiyenko 2 ID , Julio C. Rodríguez-Quiñonez 1 and Lars Lindner 2 1 2

*

ID

,

Facultad de Ingeniería, Universidad Autónoma de Baja California, Mexicali, Baja California 21280, Mexico; [email protected] (W.F.-F.); [email protected] (J.C.R.-Q.) Instituto de Ingeniería, Universidad Autónoma de Baja California, Mexicali, Baja California 21280, Mexico; [email protected] (M.R.-L.); [email protected] (O.S.); [email protected] (L.L.) Correspondence: [email protected]

Received: 9 March 2018; Accepted: 20 April 2018; Published: 24 May 2018







Abstract: The present paper describes the experimentation in a controlled environment and a real environment using different photosensors, such as infrared light emitting diode (IRLED-as receiver), photodiode, light dependent resistor (LDR), and blue LED for the purpose of selecting those devices, which can be employed in adverse conditions, such as sunlight or artificial sources. The experiments that are described in this paper confirmed that the blue LED and phototransistor could be used as a photosensor of an Optical Scanning System (OSS), because they were less sensitive to sunlight radiation. Moreover, they are appropriate as reference sources that are selected for the experiment (blue LED flashlight and light bulb). The best experimental results that were obtained contained a digital filter that was applied to the output of the photosensor, which reduced the standard deviation for the best case for the phototransistor LED from 100.26 to 0.15. For the best case, using the blue LED, the standard deviation was reduced from 86.08 to 0.11. Using these types of devices the cost of the Optical Scanning System can be reduced and a considerable increase in resolution and accuracy. Keywords: light emitting diode (LED-receiver); photodiode; real environment; light dependent resistor (LDR); optical scanning system (OSS)

1. Introduction One of the principal targets of an Optical Scanning System (OSS) is the noise reduction under real-life conditions. There are several possible sources of interference, which may cause problems when the system is exposed to adverse environmental conditions. The main sources of noise are sunlight, electrically or magnetically induced interferences, and electronic components, such as Op-amps that are used in the OSS to measure and amplify small signals, as the 60 Hz power line frequency, which is a source of substantial noise in many photosensors. On the other hand, when photodetectors are used to evaluate the performance and the accuracy of an optical system, it is critical to take into a consideration both the sensor and the light source. It is also important to consider that all commercial photosensors are sensitive to sunlight due to their spectral response of the sun. In practice, however, it is possible to find solutions using filters, such as optical filters, to attenuate the undesired wavelength radiation and analog filters to remove the electrical noise. In addition, to solve these kind of problems, there are methods and sophisticated techniques, such as computational statistics and digital filters, which can be used to reduce the noise from environmental sources. These methods and devices can improve the performance of optical scanning systems. It is important to mention that these methodologies are

Sensors 2018, 18, 1684; doi:10.3390/s18061684

www.mdpi.com/journal/sensors

Sensors 2018, 18, 1684

2 of 15

based on the quality of the data. The authors in [1] reviewed the different aspects of SHM in detail, such as feature extraction and statistical analysis. The OSS is based on optoelectronic sensors to capture an optical signal from a specific point in space and the main reason is because it is necessary to obtain the shape of an object or to monitor this point under observation. Devices, such as LED, phototransistor, photodiode, and light-dependent resistor (LDR) can be used as a natural filter due to their narrow electromagnetic spectrum, and which are available on the commercial market at low cost. However, choosing these types of sensors depend on the application, for example when the OSS is not influenced by other source of radiation. There are two types of techniques for remote object sensing, which are divided into active and passive methods [2]. Laser scanning systems as an active method for object scanning, which employs its own energy source [3]. The important advantage of using Laser Scanning Systems is that this system can control its own source of radiation and it is also possible to obtain measurements anytime. Passive methods detect natural energy that is reflected or emitted from the observed object. Also, the OSS can work with an incoherent source light, such as light emitting diode (LED) or coherent light, such as laser. The advantage of this method is that the cost of the overall system is cheaper than the active method when compared to the active method. Another important advantage of the passive method is that it can be used to monitor a structural performance by placing on the structure a source of radiation, such as a light bulb, LED, or Laser. Nowadays, the development of optoelectronic sensors that can be used in adverse conditions, such as sunlight and other sources of radiation, represents a challenge, particularly in the industrial applications of structural health monitoring (SHM), navigation of autonomous mobile robots, remote sensing, and any other application that requires sensing the light from an object. The principle of operation of the passive method is that the object being observed reflects ambient radiation that is emitted by the sun or a nearby source of artificial light [4]. Photodiodes and photoresistors, such as light-dependent resistor (LDR), are an example of photodetectors [5]. The use of optoelectronic devices can provide a natural filter, due that they operate within a specific range of wavelength, this is important to compare each device mentioned before in this paper because under real-life environmental conditions some sensors not working well. Photodiode and phototransistors are widely used as optoelectronic sensors for optical scanning. However, in [6,7] the LED has been used as a photosensor with satisfactory results. The photosensor mentioned before is used as receiver of a light that is reflected from the object and can be found most commonly in scanning systems for three-dimensional (3D) vision-based range scanning tasks. Paper [8] proposed a novel positioning system for indoor application, which can measure the angular position of a moving optical sensor by using visible light communication, photodiodes, and flickering infrared LEDs. Working under real-life environmental conditions, the accuracy of this can be affected by the interference of the sun, because its spectral response due to the infrared used responds to 940 nm. The infrared LED can be changed by far infrared LED, the accuracy and resolution may be better in spite of adversity of the environmental conditions that are caused by the sunlight. In [9], the photodiode has been used for spatial measurements using optical scanning in a controlled environment. There are other examples where there are applied photosensors, such as photodiodes, for example, OPT301 in [10]. However, when it comes to realizing the experiments in a real environment, the measurements are affecting the accuracy and precision of the OSS [11]. When the devices are near to visible light, it is most difficult to discriminate the signals that are generated by the sunlight and external sources. To solve this kind of issue, the photosensor uses a daylight blocking filter in order to select a specific operating wavelength. The present paper only deals with passive methods, and the principal elements that are used are photosensors, such as photodiodes, blue and infrared LEDs as photosensors, and light dependent resistors (LDR). The main goal of this paper is discriminating the optical noise that is generated in a real environment caused by the sunlight with devices that are widely available and inexpensive. Another challenge that will be considered in this works belongs to the field of the OSS, that is the electrical noise that appears mainly in digital and analog circuits, and because it is important

Sensors 2018, 18, 1684

3 of 15

conditioning the signal before to process it. In [12], were applied the digital filters, such as median filter and moving average filter, for the conditioning of the scattered light signals of the three photodiodes that were employed in that work. In addition, transimpedance amplifier can be used to amplify the signal that is sensed by the photodiode. 2. Theoretical Fundamentals of the Photosensors A photosensor converts electromagnetic radiation into a different physical form, usually electrical charge [13], which are then classified into two major categories: photoelectric sensors and thermal sensors [14]. It must be emphasized that the present work is based on photoelectric sensors mainly made of semiconductor materials such as: photodiodes, light emitting diodes (LED) and LDRs. To understand how photoelectric sensors works, it is important to understand the photoelectric effect. According to [15], the observation of the photoelectric effect occurs by incident light on a metal surface, which causes the emission of electrons from this surface. Also, the amount of the emitted electrons and their velocity can be measured. Depending on the color of the light, the electron speed can change, for example, when the color of the light is red, the electron speed is slow when compared with the blue light. Equation (1) shows the relation between the energy gap of material and frequency of the photon. hv ≥ Eg

(1)

where h is Planck’s constant, v is the frequency of the photon and Eg is the energy gap or band gap, known as energy difference between the valence band and the conduction band. By knowing v = λc , Equation (1) can be represented in term of wavelength λ. λ≤

hc . Eg

(2)

Equation (2) reveals the relationship between the wavelength and the bandgap of the optoelectronic sensor based on semiconductor. For example, the energy gap Eg of silicon and germanium is 1.12 eV and 0.66 eV, respectively. Table 1 illustrates the principal parameters of the semiconductors materials that are used as photosensors. Table 1. Spectral parameters of different types of semiconductors. Material Germanium Silicon GaAs GaAsP CdSe CdS ZnSe GaN InSb PbS

BandGap (eV) 0.66 1.12 1.43 1.9 1.70 2.42 2.7 3.4 0.17 0.41

Energy (J) 10−19

1.056 × 1.792 × 10−19 2.288 × 10−19 2.288 × 10−19 2.723 × 10−19 3.877 × 10−19 4.320 × 10−19 5.440 × 10−19 0.2723 × 10−19 0.656 × 10−19

Wavelength, λ (nm)

Color

1882 1109 868 654 729 512 460 365 7.29 3.026

Near Infrared Near Infrared Near Infrared Visible (Orange) Visible (Red) Visible (Green) Visible (Blue) Visible (Violet) FIR NIR

Source: Sze, S.M., Physics of Semiconductor Device, Wiley Interscience Publication, 1981, pp. 848–849.

3. Optical Scanning System In the optoelectronic and measurements laboratory of the Engineering Institute of Universidad Autónoma de Baja California, a method for the task of SHM by using an OSS has been developed. On the other hand, one of the principal targets to increment accuracy and resolution for this system is reducing the noise that is caused by sunlight and environmental conditions, due to those operational conditions that affect measured signals during data acquisition. In the field of SHM, the OSS plays

Sensors 2018, 18, 1684 Sensors 2018, 18, x FOR PEER REVIEW

4 of 15 4 of 15

an important role to measure structural deformation and displacements of buildings to monitor and prevent undesirable damage. Additionally, OSS has several applications such as 3D & 2Dsuch machine monitor and prevent undesirable damage.the Additionally, the OSS has several applications as 3D vision medical scanning, and itsscanning, necessaryand accuracy and resolution inand eachresolution activity inin & 2Dtechnologies, machine vision technologies, medical its necessary accuracy order measure minimal changesminimal to assess structural conditions [16].conditions The developed OSSdeveloped consists each to activity in order to measure changes to assess structural [16]. The principally of devices and elements, such as a non-rotating incoherent light emitter that is mounted on OSS consists principally of devices and elements, such as a non-rotating incoherent light emitter that ◦ the structure,on a 45 inside of mirror the OSS, a double andconvex an optoelectronic is mounted the-slopping structure,mirror a 45°-slopping inside of theconvex OSS, alens, double lens, and an sensor as a photodiode, asashown in Figure 1: optoelectronic sensor as photodiode, as shown in Figure 1:

Figure 1. Developed Scanning Aperture, this figure shows the principal elements of the scanning Figure 1. Developed Scanning Aperture, this figure shows the principal elements of the scanning systems. systems. The elements are listed as: (A) Scanning Aperture; (B) Optoelectronic sensor that used (light The elements are listed as: (A) Scanning Aperture; (B) Optoelectronic sensor that used (light emitting diode emitting diode (LED) photodiode OPT301 and light dependent resistor (LDR)); (C) Opto-interruptor (LED) photodiode OPT301 and light dependent resistor (LDR)); (C) Opto-interruptor ITR8102; (D) Slopping ITR8102; (D) Slopping mirror; and, (E) Oscilloscope displaying the signals generated by pulses of mirror; and, (E) Oscilloscope displaying the signals generated by pulses of ITR8102 (yellow) and Gaussian ITR8102 (yellow) and Gaussian signal (turquoise). signal (turquoise).

When two scanning apertures are used, the system can calculate coordinates and distances Whenthe twosource scanning apertures are used, the system can calculate coordinates distances between of light and OSS. Each peak of the Gaussian curve is related and to the angle of between the source of light and OSS. Each peak of the Gaussian curve is related to the angle of position position of the source, in this way, the angle 𝐵𝑖 for the SA #1 and the angle 𝐶𝑖 for the second SA #2 ofare theobtained. source, inWhen this way, the anglethat Bi for SA #1 and the angle SA #2the arecoordinates obtained. i for the second considering thethe distance 𝑎 between theCapertures is known, When considering that the distance a between the apertures is known, the coordinates of the object or of the object or structure to analyze can be calculated by using theorems of sines and the correlation structure analyze canthe betriangle. calculated by using theorems of sines and the correlation between the sides betweentothe sides in in the However, triangle. in the present work, the different types of optoelectronics sensor were tested in a However, in the present work, the different types of optoelectronics were tested in the a controlled environment (laboratory) and a real operational environment, sensor in order to compare controlled environment real operational environment, to compareofthe accuracy and resolution(laboratory) of the OSS.and Thea environment conditions affect in theorder measurements the accuracy and resolution of the OSS. The environment conditions affect the measurements of thescanning angles angles of each SA, the principal source of noise and the loss of resolution from the optical ofsystem each SA, the excessive principal bright sourceof ofsunlight, noise andasthe loss of resolution from optical scanning asystem is is the previously mentioned. Inthe [9,17], is presented complete the excessive bright of sunlight, as previously mentioned. In [9,17], is presented a complete description description of the device system principle of operation. According to these works, the OSS generates ofathe device signal, systemwhere principle operation. According to these works, the generates a Gaussian Gaussian the of peak of this signal is related to the center ofOSS energy. Once knowing the signal, where the peak of this signal is related to the center of energy. Once knowing the center of center of energy, the angle is measured, as follows. energy,The thedistance angle is of measured, as follows. is T2π equal to the time between M1 and M3, expressed in Equation (3). The distance of is T2π equal to the time between M1 and M3 , expressed in Equation (3). T2  M 3  M 1 (3) T2π = M3 − M1 (3) On the other hand, the time 𝑇𝛼 is equal to the distance between 𝑀1 and 𝑀2, as expressed by On the(4). other hand, the time Tα is equal to the distance between M1 and M2 , as expressed by Equation Equation (4). . (4) 2  TαT=MM MM (4) 2− 1. 1 where T2π and Tα are expressed in samples. With this consideration, the time variable can be eliminated from Equation (4) and  , thereby is calculated as follows.

Sensors 2018, 18, 1684

5 of 15

where T2π and Tα are expressed in samples. With this consideration, the time variable can be eliminated from Equation (4) and α, thereby is calculated as follows. Sensors 2018, 18, x FOR PEER REVIEW Sensors 2018, 18, x FOR PEER REVIEW

5 of 15

α=



2π · Tα 2T2π T 

5 of 15

(5)

(5)

2T  T The procedure for measuring the virtual angle step 1,   2αis shown in Figure 2. For example, in (5) the number of the rising and falling edge that is generated by the encoder in the variable M T The procedure for measuring the virtual angle α2is shown in Figure 2. For example, in step 1, the1 and M3 isnumber recorded. The measurement that that is shown in step corresponds to the timeM of1 and one M scanning of the rising and falling edge is generated by2the encoder in the variable 3 is The procedure for measuring the virtual angle α is shown in Figure 2. For example, in step 1, the cyclerecorded. and the The difference between starting pulses of motortoand peakofof the Gaussian measurement that the is shown in step 2 corresponds the time one scanning cyclesignal and by number of the rising and falling edge that is generated by the encoder in the variable M1 and M3 is the difference pulses Once of motor of determined, the Gaussian the signal by applying applying Equationsbetween (5) andthe (6),starting respectively. T2πand andpeak Tα are virtual angle α can recorded. The measurement that is shown in step 2 corresponds to the time of one scanning cycle and Equations as (5)shown and (6),inrespectively. Once T2π and TαDepending are determined, thefrequency virtual angle α can be calculated, step 3 using Equation (7). on the of the DCbe motor, the difference between the starting pulses of motor and peak of the Gaussian signal by applying calculated, as shown in step 3 using Equation (7). Depending on the frequency of the DC motor, the the virtual angles in a column vector of the the vector is 1 angle × n, and n isbeequal Equations (5) are andstored (6), respectively. Once T2π where and Tαthe arelength determined, virtual α can virtual angles are stored in a column vector the length of the vectorfrequency is 1 × n, and nthe is equal to to thecalculated, scanning frequency, toEquation step 4. where For the DC motor is as shown inaccording step 3 using (7).example, Depending onscanning the frequency of the of DC motor, the the scanning frequency, according to step 4. For example, the scanning frequency of the DCTmotor is virtual angles are stored in a column vector where the length of the vector is 1 × n, and n is equal to 20 Hz, which means that the dimension of the column vector it will be equal to (1 × 20 ) . 𝑇 20 Hz, which means that the dimension of the column vector it will be equal to (1 × 20) . the scanning frequency, according to step 4. For example, the scanning frequency of the DC motor is 20 Hz, which means that the dimension of the column vector it will be equal to (1 × 20)𝑇 .

Figure 2. Procedure of measuringofofthe thevirtual virtual angle angle between andand the the reference Figure 2. Procedure of measuring betweenthe thephotosensor photosensor reference source (light bulb/LED flashlight). source (light2.bulb/LED Figure Procedure flashlight). of measuring of the virtual angle between the photosensor and the reference source (light bulb/LED flashlight).

The photosensors, such as photodiode, phototransistor, LDR, and Infrared LED was capable to

The such photodiode, phototransistor, LDR, and Infrared LED was capable to sensephotosensors, the light bulb that is as illustrated in Figure 3. The photosensors, such as photodiode, phototransistor, LDR, and Infrared LED was capable to sensesense the light bulb that is illustrated in Figure 3. the light bulb that is illustrated in Figure 3.

Figure 3. This figure shows a light bulb that is used as a source of radiation and the output of the Optical Scanning System (OSS) displayed by an oscilloscope. Figure 3. This figure shows a light bulb that is used as a source of radiation and the output of the Figure 3. This figure shows a light bulb that is used as a source of radiation and the output of the Optical Scanning System (OSS) displayed by an oscilloscope.

The experimental set-up the optoelectronic sensor evaluation that was used in this paper is Optical Scanning System (OSS)for displayed by an oscilloscope. presented in Figure 4. The main elements are used with environmental conditions are: a reference The experimental set-up for the optoelectronic sensor evaluation that was used in this paper is presented in Figure 4. The main elements are used with environmental conditions are: a reference

Sensors 2018, 18, 1684

6 of 15

The experimental set-up for the optoelectronic sensor evaluation that was used in this paper is 6 of 15 presented in Figure 4. The main elements are used with environmental conditions are: a reference source, OSS, data acquisition (DAQ’s system), and a computer to process the signal that was captured. source, OSS, data acquisition (DAQ’s system), and a computer to process the signal that was This figure illustrates the OSS system at 90◦ with respect to the source of light, in this example, captured. This figure illustrates the OSS system at 90° with respect to the source of light, in this a flashlight is used. However, it used a blue lamp as energy source and a LED as a photosensor example, a flashlight is used. However, it used a blue lamp as energy source and a LED as a with a maximum response at 420 nm wavelength. The reason for choosing a light bulb as energy photosensor with a maximum response at 420 nm wavelength. The reason for choosing a light bulb source for the testing process is due to its wide spectral range to radiate energy to the phototransistor, as energy source for the testing process is due to its wide spectral range to radiate energy to the LED, and photodiode that were used in this work. For example, the best response of the infrared phototransistor, LED, and photodiode that were used in this work. For example, the best response of LED and phototransistor is around 940 nm and the spectral response of the used LDR is around the infrared LED and phototransistor is around 940 nm and the spectral response of the used LDR is 600 nm wavelength. around 600 nm wavelength. Sensors 2018, 18, x FOR PEER REVIEW

Figure figure shows Figure4.This 4. This figure showsa aflashlight flashlightused useda asource sourceofofradiation radiationand andthe theoutput outputofofthe theOSS OSSdisplayed displayed acquired acquiredby bydata dataacquisition acquisition(DAQ) (DAQ)and anddisplayed displayedatatthe thecomputer. computer.

4.4.Results Results This Thissection sectiondescribes describesthe theexperiments experimentsthat thatwere weredone doneininthe thelaboratory laboratoryand andthe thereal realoperational operational environment. The data in Tables 2, 4, 6, and 8 are from experiments that are realized only in environment. The data in Tables 2, 4, 6, and 8 are from experiments that are realized only in laboratory, laboratory, where the samples were the taken from the OSS, first, as following: first,had each sample had where the samples were taken from OSS, as following: each sample a duration of 10as. duration of 10 s. The parameters and type of input of the DAQ system were sampled using a 50 KHz The parameters and type of input of the DAQ system were sampled using a 50 KHz sampling frequency sampling frequency and two analog input channels were used for data acquisition. The motor and two analog input channels were used for data acquisition. The motor frequency of the OSS was frequency the OSS was 20the Hz.experiment, In order to realize the experiment,were threetaken, configurations were taken, 20 Hz. Inoforder to realize three configurations such as changing the such as changing the position of the reference source. position of the reference source. The Thefirst firstconfiguration configurationofofthe theexperiment experimentwas wascalled calledPosition Position90°, 90◦ ,and andthe thesecond secondand andthird third ◦ ◦ ◦ configuration of the experiment were 91° and 92°, respectively. Position 90° means that blue flashlight configuration of the experiment were 91 and 92 , respectively. Position 90 means that blue flashlight ◦ with was respect to the position of the at a distance of 1 m. was taken wassituated situatedatat90° 90with respect to the position of OSS the OSS at a distance ofEach 1 m. sample Each sample was during 10 s in10duration, for example in sample 1, the standard deviation mean µµ1 were taken during s in duration, for example in sample 1, the standard deviationσ1σand and mean were 1 1 calculated. calculated.Also, Also,ititshould shouldbe beemphasized emphasizedthat thateach eachsample samplecorresponds correspondstotothe theangles anglescalculated calculatedby by the theOSS OSSininaacontrolled controlledenvironment. environment.All Alldata dataininthe theTables Tables2–5 2–5were werecalculated calculatedby bydeveloping developingan an algorithm algorithmininMatLab MatLaband andcomparing comparingthe thereal realangle angleofofthe thesource source(90°, (90◦ ,91°, 91◦ ,and and92°) 92◦ )with withrespect respectofof the angle that was detected by the algorithm of SA by returning one column vector each number the angle that was detected by the algorithm of SA by returning one column vector each numberofof sample. ofof time forfor allall of of thethe configurations was 300300 s, with 100100 s corresponding to sample.The Thetotal totalamount amount time configurations was s, with s corresponding each configuration. TheThe column vector, called 𝑆𝑎𝑚𝑝𝑙𝑒1 , has = 217 elements, if we multiply to each configuration. column vector, called Sample1, has𝑚m𝐷𝐴𝑄 DAQ = 217 elements, if we multiply the number of pulses of the motor (20 pulses/seconds) by the time that the number of pulses of the motor (20 pulses/seconds) by the time thatwas wassampled sampledby byDAQ DAQ(10 (10s),s), we that the thenumber numberofofangles angles calculated 𝑚𝑡ℎ𝑒𝑜𝑟𝑖𝑐𝑎𝑙 = 220, is not toequal to elements the 217 we note note that calculated is mistheorical = 220, whichwhich is not equal the 217 elements calculated. This is due to the motor frequency was not controlled. Each sample is saved in Equation (6), as following. 𝑆𝑎𝑚𝑝𝑙𝑒1 = [90.537 90.494 90.581 … 𝑆𝑎𝑚𝑝𝑙𝑒1(𝑚−2) 𝑆𝑎𝑚𝑝𝑙𝑒1(𝑚−1) 𝑆𝑎𝑚𝑝𝑙𝑒1(𝑚) ]𝑇

(6)

Sensors 2018, 18, 1684

7 of 15

calculated. This is due to the motor frequency was not controlled. Each sample is saved in Equation (6), as following. Sensors 2018, 18, x FORh PEER REVIEW

Sample1 =

90.537

90.494

90.581 . . . Sample1(m−2)

Sample1(m−1)

Sample1(m)

i T 7 of 15 (6)

Table 2 shows the experiment with blue LED in a controlled environment. For this experiment, Table with 2 shows experiment with LED in a controlled environment. this 10 experiment, the results bluethe LED as a sensor canblue be appreciated, each sample was taken For during s and the the results with blue LED as a sensor can be appreciated, each sample was taken during 10 andwas the motor frequency of the OSS was 20 Hz. In position 90°, the minimum standard deviation sthat ◦ motor frequency thesamples OSS was 20 σHz. In position 90 calculated , the minimum deviation that was obtained from the of total was = 0.1°, which was from astandard row vector with 200 angles ◦ , which was calculated from a row vector with 200 angles obtained from the total samples was σ = 0.1 that were captured by the OSS during 10 s sample time. For the position 91° and 92°, the minimum that weredeviation captured obtained by the OSS during 10 s0.13, sample time. For the position 91◦ and 92◦ , the minimum standard was 0.08 and respectively. standard deviation obtained andwas 0.13,confronted respectively. On the other hand, thewas blue0.08 LED to the adverse conditions caused by the On the other hand, the blue LED was confronted to the adverse conditions caused the sunlight. The signal captured from the blue LED was smoothed using a Savitzky-Golay filterby(SG) sunlight. The signal capturedoffrom blue LED was Savitzky-Golay filter different (SG) [18]. [18]. Recalling the purpose thisthe paper, which is smoothed defined byusing the acomparison between Recalling the purpose of this paper, which is defined by the comparison between different photosensors photosensors In Figure 5 the signal generated by the blue LED is illustrated, note that signal in red In Figure 5 the signal by thewithout blue LEDa is illustrated, noteinthat signal in red color is the the output color is the output of generated the OSS system filter. The signal green color represents signal of the OSS system without a filter. The signal in green color represents the signal filtered using the filtered using the SG filter. SG filter. Table 2. Experiment in a laboratory using one blue LED as light source and one LED used as photosensor. Table 2. Experiment in a laboratory using one blue LED as light source and one LED used as photosensor.

Position 90° ◦ σ1 Position µ1 90 # Samples 0.13 σ1 89.54µ1 1 0.12 0.13 89.76 89.54 2 0.16 0.12 89.72 89.76 3 0.13 0.16 89.97 89.72 4 0.13 89.97 0.15 0.15 89.79 5 89.79 6 0.17 0.17 90.11 90.11 7 0.13 0.13 90.16 90.16 8 0.14 0.14 89.93 89.93 9 0.15 89.76 0.15 89.76 10 0.10 89.56 0.10 89.56

# Samples 1 2 3 4 5 6 7 8 9 10

Position 91°

Position 91◦µ2 σ 2 σ2 0.12 0.12 0.12 0.12 0.15 0.15 0.28 0.28 0.19 0.19 0.15 0.15 0.17 0.17 0.17 0.17 0.13 0.13 0.08 0.08

µ90.95 σ3 2

90.86 90.95 90.86 91.03 91.03 90.56 90.56 90.53 90.53 90.78 90.78 90.61 90.61 90.59 90.59 90.83 90.83 91.04 91.04

Position 92° µ3 µ3 0.28 92.01 0.25 92.10 92.01 92.10 0.23 92.07 92.07 0.19 92.19 92.19 0.23 92.40 92.40 0.13 92.44 92.44 91.90 0.17 91.90 91.60 0.29 91.60 91.69 0.13 91.69 91.69 0.13 91.69

◦ Positionσ92 3

0.28 0.25 0.23 0.19 0.23 0.13 0.17 0.29 0.13 0.13

Figure Figure5.5.Signal Signalsmoothed smoothed(green) (green)by byusing usingdigital digitalfilter filter by by using usingaa blue blueLED LED as as aa photosensor. photosensor.

In Figure 6a, the signal that was displayed by the oscilloscope is presented. Note that, in this case, the flashlight was turned off in order to compare when this is turned on. The voltage peak without flashlight was 1.52 V, while using a flashlight it was turned on the OSS could measure a voltage peak of 5.36 V, see Figure 6b. Table 6 summarizes statistical data from the experiment using a blue LED as a photosensor and

Sensors 2018, 18, 1684

8 of 15

In Figure 6a, the signal that was displayed by the oscilloscope is presented. Note that, in this case, the flashlight was turned off in order to compare when this is turned on. The voltage peak without flashlight was 1.52 V, while using a flashlight it was turned on the OSS could measure a voltage peak of 5.36 V, see Figure 6b. Table 6 summarizes statistical data from the experiment using a blue LED as a photosensor and a blue LED flashlight as an energy source. The statistical data were calculated, as follows. First, it was taken a sample of 10 s for each position of the reference source, and then the Sensors 2018, 18,with x FOR duration PEER REVIEW 8 ofnumber 15 of peak sample of the Gaussian signal generated by the blue LED was detected. These numbers of of peak sample of theusing Gaussian signal generated the blue the LEDnumber was detected. These of numbers of and samples were recorded a column vector. Byby knowing of samples the rising samples recorded using from a column vector. By knowing the number samples of the rising placed and falling edge were that was provided the pulses that were generated byofthe opto-interrupter in falling edge that was provided from the pulses that were generated by the opto-interrupter placed in the motor, these samples are recorded in another column vector. With these column vectors, the angles the motor, these samples are recorded in another column vector. With these column vectors, the are calculated, due to the period are known by the numbers of samples when the pulses are measured. angles are calculated, due to the period are known by the numbers of samples when the pulses are Since the full period represents a cycle of 360◦ , the next Equation (7), can be used: measured. Since the full period represents a cycle of 360°, the next Equation (7), can be used:

 α =

Gaussian _ peak _(#. sample)) Gaussian_peak_ (#.sample × ×360 360 Signal _ period (#. sample Signal_period(#.sample )

(7)

(7)

Figure 6. The output signal from a blue LED as a photosensor are displayed by oscilloscope. (a) These

Figure 6. The output signal from a blue LED as a photosensor are displayed by oscilloscope. (a) These figures illustrate the output from a LED as a photosensor exposed to environmental conditions figures illustrate the output from a LED as a photosensor exposed to environmental conditions without without source of radiation; (b) These figures show the output from the LED used as photosensor by source of radiation; (b) These figures show the output from the LED used as photosensor by using a using a flashlight as a source of radiation in sunlight. flashlight as a source of radiation in sunlight.

It must be noted that the motor frequency was not controlled, for this reason in Table 3 at 90°, 91°, and be 92°,noted the number ofmotor samples to calculate the frequency of motor is decreased. According to◦ , 91◦ , It must that the frequency was not controlled, for this reason in Table 3 at 90 this table, the difference is less than 0.06°, whenof considering average calculated forto the and the 92◦data , thefrom number of samples to calculate the frequency motor is the decreased. According the angles selected for the experiment such as 90° and 91°, which the results were 89.94°, 91.97° versus ◦ data from this table, the difference is less than 0.06 , when considering the average calculated for the the correct angles 90° and 92°. However, the difference between the correct angles 91° and the virtual angles selected for the experiment such as 90◦ and 91◦ , which the results were 89.94◦ , 91.97◦ versus angle calculated 91.53°, resulting in 0.53°. It is clear that the variance decreases considerably when ◦ . However, the difference between the correct angles 91◦ and the virtual the correct angles 90◦ and the signal is filtered in 92 order to smooth the data. For example, for the position 90°, the variance ◦ ◦ anglecalculated calculated 91.53 , resulting 0.53 is clear the variance decreases considerably when the was 8054.03 without in filter and. It 0.02 using that a filter. signal is filtered in order to smooth the data. For example, forused the position 90◦ , the variance calculated In Table 4, the results with Infrared LED as a sensor in laboratory conditions can be each sample using a 10 s duration and the motor frequency of the OSS was 20 Hz. Notice was appreciated, 8054.03 without filter and 0.02 using a filter. that in this experiment, the resolution of the system has decreased in comparison to the experiment with blue LED due to 2.4 V peak of the infrared LED versus 4.8 V peak of the blue LED. In position 90°, the minimum standard deviation that was obtained from the total samples was σ = 0.17°, which was calculated from a row vector with 200 angles that were captured by OSS with a duration of 10 s

Sensors 2018, 18, 1684

9 of 15

In Table 4, the results with Infrared LED as a sensor used in laboratory conditions can be Sensors 2018, 18,each x FOR PEER REVIEW 15 appreciated, sample using a 10 s duration and the motor frequency of the OSS was 20 Hz. Notice9 of that in this experiment, the resolution of the system has decreased in comparison to the experiment with each LED sample. position 91°ofand the minimum standard deviation was obtained was 0.14° blue dueFor to 2.4 V peak the92°, infrared LED versus 4.8 V peak of thethat blue LED. In position 90◦ , andminimum 0.17, respectively. the standard deviation that was obtained from the total samples was σ = 0.17◦ , which was calculated from a row vector with 200 angles that were captured by OSS with a duration of 10 s each Table 3. Statistical data from blue LED used as photosensor in a real environment. sample. For position 91◦ and 92◦ , the minimum standard deviation that was obtained was 0.14◦ and 0.17◦ , respectively. Position 90° Position 91° Position 92° # Statistical Data Without Smoothed Without Smoothed Without Smoothed Table 3. StatisticalFilter data from blue photosensor a real environment. by LED SG used as Filter byinSG Filter by SG Average 186.31 89.94 185.62 91.53 188.55 91.97 Position 90◦ Position 91◦ Position 92◦ # Statistical Data Standard deviation 89.74 0.16Without Filter 86.08Smoothed by 0.11 87.13 Smoothed 0.11by SG Without Filter Smoothed by SG SG Without Filter Variance 8054.03 0.02 7410.63 0.013 7593.19 0.013 Average 186.31 89.94 185.62 91.53 188.55 91.97 Standard deviation 0.11 172 87.13 0.11 Samples 89.74 1135 0.16 174 86.08 1133 1068 171 Variance 8054.03 0.02 7410.63 0.013 7593.19 0.013 Min. Value 1135 0.35 174 89.59 1133 0.35 0.35 92.59 Samples 172 91.04 1068 171 Min. Value 0.35 89.59 0.35 91.04 0.35 92.59 Max. Value 359.12 90.34 359.14 91.88 359.16 93.25 Max. Value 359.12 90.34 359.14 91.88 359.16 93.25 Table 4. Experiment in a laboratory with bulb source light and LED infrared used as photosensor. Table 4. Experiment in a laboratory with bulb source light and LED infrared used as photosensor.

Position 90° # Samples Position 90◦ σ1 µ1 # Samples σ µ 1 1 1 0.43 90.56 0.28 90.56 90.55 12 0.43 23 0.28 0.44 90.55 90.26 3 0.44 90.26 4 0.28 90.46 4 0.28 90.46 5 0.25 90.40 90.40 5 0.25 0.19 90.38 90.38 66 0.19 77 0.38 0.38 90.31 90.31 88 0.40 0.40 90.43 90.43 9 0.17 90.33 9 0.17 90.32 90.33 10 0.23 10 0.23 90.32

Position 91° Position 912◦ σ2 µ σ2 µ2 0.35 91.02 0.33 90.92 0.35 91.02 0.33 90.92 0.31 90.76 0.31 90.76 0.25 91.09 0.25 91.09 0.48 90.44 0.48 90.44 0.34 90.35 0.34 90.35 0.39 90.16 0.39 90.16 0.14 90.46 0.14 90.46 0.20 90.41 0.20 90.41 0.28 90.62 0.28 90.62

Position 92° ◦ Position σ 3 µ92 3 σ3 µ3 0.22 91.27 0.30 0.22 91.54 91.27 0.30 91.06 91.54 0.18 0.18 91.06 0.34 91.55 0.34 91.55 0.35 0.35 91.70 91.70 0.28 0.28 91.51 91.51 0.17 91.12 91.12 0.17 0.28 91.39 91.39 0.28 0.43 91.41 0.43 91.41 0.30 91.44 0.30 91.44

In can be be appreciated appreciated that that the the infrared infrared LED LED was was saturated saturated by by sunlight, sunlight, In the the following following Figure Figure 7, 7, it it can thereby this experiment could not be realized. However, in laboratory conditions, the infrared thereby this experiment could not be realized. However, in laboratory conditions, the infrared LED LED works works without without problems. problems.

Figure 7. Signal captured using an infrared LED as photosensor of the OSS in the real environment. Figure 7. Signal captured using an infrared LED as photosensor of the OSS in the real environment.

Note in Table 5, that the results in term of average and standard deviation were taken using a LDR as a photosensor with sampling time 𝑇𝑆 = 10 s and the motor frequency of 𝑓𝑀 = 20 Hz. The resolution of the OSS is about 1°, according to the results that are shown in this table. Figure 8 shows the output from the LDR it can be appreciated, that the rising and falling time of this sensor is different. However, the results that were obtained with this photosensor can be considered in order

Sensors 2018, 18, 1684

10 of 15

Note in Table 5, that the results in term of average and standard deviation were taken using a LDR as a photosensor with sampling time TS = 10 s and the motor frequency of f M = 20 Hz. The resolution of the OSS is about 1◦ , according to the results that are shown in this table. Figure 8 shows the output from the LDR it can be appreciated, that the rising and falling time of this sensor is different. However, the results that were obtained with this photosensor can be considered in order to develop an application, where the response time is dispensable. In the position 90◦ the minimum Sensors 2018, 18, x FOR PEER REVIEW 10 of 15 standard deviation obtained was σ = 0.14◦ calculated from a row vector with 200 angles captured by ◦ and to develop an application, where response is dispensable. In the position 90° the minimum OSS during 10 s each sample. For thethe position 91time 92◦ , the minimum standard deviation obtained ◦ ◦ standard deviation obtained was σ = 0.14° calculated from a row vector with 200 angles captured by was 0.20 and 0.26 , respectively. during 10 s8aeach sample. the position 91°that and 92°, minimum standard deviation obtained TheOSS next Figure shows the For experimentation wasthe carried out in a controlled environment, such was 0.20° and 0.26, respectively. as laboratory, and the Figure 8b illustrates the output that was generated by LDR with the interference of The next Figure 8a shows the experimentation that was carried out in a controlled environment, the sunlight. Note that LDR sense8bthe reference thegenerated exposurebytoLDR the sunlight. such as laboratory, andcannot the Figure illustrates thesource outputdue thatto was with the Table 6 shows the results of the experiments using a phototransistor as photosensor and interference of the sunlight. Note that LDR cannot sense the reference source due to the exposure to a bulb sunlight.source, operating in a controlled environment. The results in terms of standard light asthe reference Tableposition 6 shows the the experiments as photosensor bulb hand, deviation from 90◦results , 91◦ ,ofand 92◦ are 0.2,using 0.14,a phototransistor and 0.32, respectively. Onand theaorder light as reference source, operating in ◦a controlled environment. The results in terms of standard ◦ ◦ the best results for the mean value are 90 , 91.13 , and 92.12 , respectively. deviation from position 90°, 91°, and 92° are 0.2, 0.14, and 0.32, respectively. On the order hand, the Once the parameters of standard deviation and mean are established from the experiment that best results for the mean value are 90°, 91.13°, and 92.12°, respectively. was realizedOnce in the the next step was to experiment in a place exposure to environmental thelaboratory, parameters of standard deviation and mean are established from the experiment that agents, was suchrealized temperature and dust, with sunlight. Figure 9 exposure illustrates the experiment that in the laboratory, theespecially next step was to experiment in a place to environmental agents,in such temperature andenvironment dust, especially with sunlight. Figure 9 illustrates the experiment that was realized a real operation using a phototransistor as photosensors. According to was realized in a real operation environment using a phototransistor as photosensors. According to Figure 9a, the OSS is detecting the reflection of the sunlight on the surfaces from the environment. Figure 9a, the OSS is detecting the reflection of the sunlight on the surfaces from the environment. For this case, the bulb light was turned off in order to make it possible to distinguish between the For this case, the bulb light was turned off in order to make it possible to distinguish between the different external sources of light that were bythe thesunlight. sunlight. Figure 9b shows the that signal that different external sources of light that werecaused caused by Figure 9b shows the signal was detected by thebyOSS when thethe bulb light on. was detected the OSS when bulb lightwas was turned turned on.

(a)

(b)

Figure 8. Signal captured by using a LDR as photosensor of the OSS, in (a), the output was taken from

Figure 8. Signal captured by using a LDR as photosensor of the OSS, in (a), the output was taken from a controlled environment such as laboratory. In the (b), is illustrated the output that was generated a controlled environment in adverse conditions.such as laboratory. In the (b), is illustrated the output that was generated in adverse conditions. Table 5. Experiment in a laboratory with bulb source light and LDR used as photosensor.

Table 5. Experiment in a laboratory with90° bulbPosition source light LDR92° used as photosensor. Position 91° and Position # Samples σ1 µ1 σ2 µ2 σ3 µ3 Position 90◦ Position 91◦ 0.26 Position 92◦ 1 0.14 89.61 0.26 90.62 91.74 # Samples σ0.35 µ1 σ2 µ2 σ3 91.48µ3 2 89.90 0.20 90.76 0.34 1 3 0.24 89.66 0.55 91.28 0.25 1 0.14 89.61 0.26 90.62 0.26 91.56 91.74 4 0.35 89.66 0.30 90.64 0.25 2 0.35 89.90 0.20 90.76 0.34 91.56 91.48 3 5 0.24 0.55 90.63 91.28 0.26 0.25 91.61 91.56 0.46 89.66 90.02 0.42 4 6 0.35 0.30 90.77 90.64 0.26 0.25 91.55 91.56 0.16 89.66 89.95 0.41 5 7 0.46 0.42 90.89 90.63 0.35 0.26 91.66 91.61 0.19 90.02 89.91 0.33 6 0.16 89.95 0.41 90.77 0.26 91.55 8 0.25 89.43 0.43 90.88 0.51 91.90 7 0.19 89.91 0.33 90.89 0.35 91.66 0.17 89.43 89.78 0.33 8 9 0.25 0.43 90.73 90.88 0.30 0.51 91.81 91.90 0.33 89.78 89.71 0.29 9 10 0.17 0.33 90.80 90.73 0.35 0.30 91.59 91.81 10

0.33

89.71

0.29

90.80

0.35

91.59

Sensors FOR PEER REVIEW Sensors 2018, 2018, 18, 18, x1684

11 11 of of 15 15

Table 6. Experiment in a laboratory with bulb source light and Phototransistor used as photosensor. Table 6. Experiment in a laboratory with bulb source light and Phototransistor used as photosensor.

# Samples # Samples

1 2 1 23 34 45 56 6 77 88 99 10 10

Position 90° σ1 µ1◦ Position 90 0.30 90.18 σ1 µ1 0.28 90.00 0.30 90.18 0.20 90.00 90.25 0.28 0.20 90.25 90.33 0.20 0.20 90.33 0.29 90.35 0.29 0.29 90.35 90.10 0.29 90.10 0.41 89.62 0.41 89.62 0.34 89.80 89.80 0.34 0.46 0.46 89.93 89.93 0.48 0.48 90.23 90.23

Position 91° σ2 µ Position 912 ◦ 0.34 91.16 σ2 µ2 0.23 91.33 0.34 91.16 0.27 0.23 91.30 91.33 0.14 0.27 91.33 91.30 0.14 91.33 0.28 91.13 0.28 91.13 0.34 91.22 0.34 91.22 0.44 0.44 90.90 90.90 0.42 0.42 91.13 91.13 0.56 91.36 91.36 0.56 0.32 91.36 91.36 0.32

Position 92° σ 3 µ92 3 ◦ Position 0.48 σ3 92.44 µ3 0.32 92.12 0.48 92.44 0.33 0.32 92.33 92.12 0.33 0.33 92.41 92.33 0.33 92.41 0.43 92.18 0.43 92.18 0.29 92.61 0.29 92.61 0.51 0.51 92.66 92.66 0.59 0.59 92.67 92.67 0.43 92.58 92.58 0.43 0.37 92.41 92.41 0.37

Figure 9. 9. Signal Signal captured captured by by using using aa phototransistor. phototransistor. (a) (a) The The phototransistor phototransistor detects detects peaks peaks of of voltage voltage Figure caused by environmental factors (external agents such as temperature and other source of radiation); caused by environmental factors (external agents such as temperature and other source of radiation); (b) In In spite spite of sun's radiant (b) of the the sun's radiant energy, energy, the the system system OSS OSS could could detect detect the thebulb bulb light. light.

After the signal was detected detected by by inspection, inspection,as asmentioned mentionedbefore, before,the thesignal signalwas wasacquired acquiredusing usinga aDAQ DAQ with kHz sampling a motor DC motor angular frequency of 20 Hz. The reference with 5050 kHz sampling raterate andand a DC withwith angular frequency of 20 Hz. The reference source was placed at 90◦ , 91 , and 92◦and with to the OSS. 10 reveals that thethat Gaussian signal source was placed at ◦90°, 91°, 92°respect with respect to theFigure OSS. Figure 10 reveals the Gaussian that was generated by the phototransistor was clipped at 5 V,off dueatto5the signal that was generated by the phototransistor wasoff clipped V, gain due of to this the photosensor gain of this exceeds the power supply. photosensor exceeds the power supply.

Sensors 2018, 18, 1684

12 of 15

Sensors 2018, 18, x FOR PEER REVIEW

12 of 15

Figure Figure 10. 10. Signal Signalcaptured capturedusing usingaaphototransistor phototransistor as as photosensor photosensor in in the the OSS. OSS.

Table 7 shows the statistical data that were calculated from phototransistor and provides an Table 7 shows the statistical data that were calculated from phototransistor and provides an estimation of when the angles between the OSS and light of source have been calculated. There are estimation of when the angles between the OSS and light of source have been calculated. There are two columns for each angle, where the first column contains data processed without a filter, and the two columns for each angle, where the first column contains data processed without a filter, and the second column contains data smoothed using the Savitzky-Golay filter. The difference is considerable second column contains data smoothed using the Savitzky-Golay filter. The difference is considerable when the signal is filtered. For example, at the angle 90°,◦ the system calculated an angle of 189.13°.◦ when the signal is filtered. For example, at the angle 90 , the system calculated an angle of 189.13 . However, the real angle is 90°, resulting a difference of −99.13°, on the other hand, the difference is However, the real angle is 90◦ , resulting a difference of −99.13◦ , on the other hand, the difference is reduced to 0.39°◦ when the signal is smoothed by the digital filter. reduced to 0.39 when the signal is smoothed by the digital filter. Table 7. Statistical data from phototransistor used as photosensor in a real environment. Table 7. Statistical data from phototransistor used as photosensor in a real environment.

Position 90° Position 91° Position 92° Position 91◦ Position 92◦ Without Smoothed Without Smoothed Without Smoothed Without Filter Smoothed by SG Without Filter Smoothed by SG Without Filter Smoothed by SG Filter by SG Filter by SG Filter by SG Average 189.13 89.61 189.81 90.72 193.72 91.78 Standar deviation 100.26 189.13 0.15 105.20 0.19 Average 89.61 101.49 189.81 0.22 90.72 193.72 91.78 Variance 10052.19 0.02 10301.53 0.048 11068.63 0.039 Standar deviation 1081 100.26 200 0.15 971 101.49 200 0.22 0.19 Samples 823105.20 196 Min. Value 0.39 10052.19 89.19 0.38 91.54 Variance 0.02 0.40 10301.53 90.32 0.048 11068.63 0.039 Max. Value 358.97 90.39 358.96 91.49 359.01 92.36 Samples 1081 200 971 200 823 196 Min. Value 0.39 89.19 0.40 90.32 0.38 91.54 As it can be seen from358.97 Figure 11, it was to detect91.49 the signal in the real operational Max. Value 90.39not possible 358.96 359.01 92.36 environment generated reference source (bulb light) by using the photodiode OPT301 as a photosensor, due As to the interference with sunlight. Figure illustrates the experiment was operational realized in it can be seen from Figure 11, it was not 11 possible to detect the signal inthat the real sunlight and the output reference from OPT301 that(bulb was displayed the oscilloscope. Figure 11a shows environment generated source light) by by using the photodiode OPT301 as a the OSS, the source of light (light bulb), and Figure 11b illustrates the signal that was displayed by the photosensor, due to the interference with sunlight. Figure 11 illustrates the experiment that was oscilloscope. It must be noted that the output in yellow color is caused by reflections of the sunlight. realized in sunlight and the output from OPT301 that was displayed by the oscilloscope. Figure 11a The results fromthe the photodiode in laboratory are11b shown in Table shows the OSS, source of lightOPT301 (light bulb), and Figure illustrates the8.signal that was displayed ## Statistical Statistical Data Data

Position 90◦

by the oscilloscope. It must be noted that the output in yellow color is caused by reflections of the sunlight. The results from the photodiode OPT301 in laboratory are shown in Table 8.

Sensors 2018, 18, Sensors 2018, 18, 1684 x FOR PEER REVIEW

13 of of 15 15 13

Table 8. Experiment in a laboratory with bulb source light and photodiode OP301 used as photosensor. Table 8. Experiment in a laboratory with bulb source light and photodiode OP301 used as photosensor.

# Samples # Samples

1 12 23 34 4 5 5 66 77 88 99 10 10

Position 90° Position 91° Position 912◦ σ2 µ σ µ1 σ2 µ2 0.23 90.40 0.18 91.81 1 0.21 90.46 0.14 91.57 0.23 90.40 0.18 91.81 0.21 0.14 91.68 91.57 0.18 90.46 90.46 0.19 0.18 0.19 91.49 91.68 0.29 90.46 90.34 0.23 0.29 90.34 0.23 91.49 0.17 90.29 0.27 91.65 0.17 90.29 0.27 91.65 0.26 90.27 90.27 0.23 0.26 0.23 91.63 91.63 0.20 90.44 90.44 0.28 0.20 0.28 91.89 91.89 0.20 90.45 0.20 91.52 0.20 90.45 0.20 91.52 0.16 90.37 0.12 91.34 0.16 90.37 0.12 91.34 0.25 90.49 0.12 91.42 0.25 90.49 0.12 91.42 ◦ Position 90 σ1 µ1

Position 92° ◦ Position σ 3 µ92 3 σ3 92.65 µ3 0.20 0.21 92.92 0.20 92.65 0.21 92.80 92.92 0.19 0.19 92.11 92.80 0.29 0.29 92.11 0.14 92.84 0.14 92.84 0.13 0.13 92.81 92.81 0.14 0.14 92.56 92.56 0.12 92.73 0.12 92.73 0.16 92.78 0.16 92.78 0.11 92.62 0.11 92.62

Figure 11. Data acquisition by using OPT301 in a real operation: (a) This figure shows the OSS system Figure 11. Data acquisition by using OPT301 in a real operation: (a) This figure shows the OSS system and bulb light lightas asenergy energysource sourceand andOPT OPT 301 a photosensor in the operational environment; and bulb 301 as as a photosensor in the realreal operational environment; and, and, (b) Oscilloscope display the signal generated by pulses of ITR8102 (turquoise) and Gaussian (b) Oscilloscope display the signal generated by pulses of ITR8102 (turquoise) and Gaussian signal signal (yellow) from sunlight. (yellow) from sunlight.

5. Conclusions 5. Conclusions The comparison between different types of photosensors has been presented in order to select a The comparison between different types of photosensors has been presented in order to select photosensor, which can be employed in a hostile environment where naturals and external light a photosensor, which can be employed in a hostile environment where naturals and external light sources are presented. The photosensors that can detect the reference light source exposed to sunlight sources are presented. The photosensors that can detect the reference light source exposed to sunlight conditions, are visualized in the Table 9. The results with the phototransistor and LED show that it is conditions, are visualized in the Table 9. The results with the phototransistor and LED show that it is possible to discriminate the reflection of the sunlight with the support of digital filters, such as FIR possible to discriminate the reflection of the sunlight with the support of digital filters, such as FIR filters. Under sunlight conditions, using the blue LED gives satisfactory results, including a standard filters. Under sunlight conditions, using the blue LED gives satisfactory results, including a standard deviation of 0.16 and an average of 89.94 at 90° versus a standard deviation of 0.138 and an average deviation of 0.16 and an average of 89.94 at 90◦ versus a standard deviation of 0.138 and an average of of 89.83° for the same angle of reference. The results from the blue LED as a sensor and 89.83◦ for the same angle of reference. The results from the blue LED as a sensor and phototransistor phototransistor that was obtained from both the laboratory and outdoors can be considered to be that was obtained from both the laboratory and outdoors can be considered to be applied by OSS with applied by OSS with adverse conditions. It is important to mention that the error in angle adverse conditions. It is important to mention that the error in angle measurements for the blue LED measurements for the blue LED in sunlight conditions were better (0.06°) than the experiments that in sunlight conditions were better (0.06◦ ) than the experiments that were carried out under laboratory were carried out◦ under laboratory conditions (0.17°). The reason for this difference is due to the signal conditions (0.17 ). The reason for this difference is due to the signal sensed in the adverse environment sensed in the adverse environment had to be smoothed by digital filter. had to be smoothed by digital filter. All of the photosensors were placed at a distance of 1 m with respect to the light source. All of the photosensors were placed at a distance of 1 m with respect to the light source. However, in the case of the gain of the phototransistor at this distance, the output was clipped off However, in the case of the gain of the phototransistor at this distance, the output was clipped because the voltage gain of this photosensor exceeded the power supply. According to the datasheet, off because the voltage gain of this photosensor exceeded the power supply. According to the the peak response wavelength is at 940 nm, while considering that the literature part of this region is datasheet, the peak response wavelength is at 940 nm, while considering that the literature part blocked by atmospheric water vapor that is centered at this wavelength. Furthermore, the of this region is blocked by atmospheric water vapor that is centered at this wavelength. Furthermore, phototransistor contains a daylight blocking filter in order to reduce the sunlight radiation. The use of the blue LED in environmental conditions resulted that this device can attenuate the radiation that

Sensors 2018, 18, 1684

14 of 15

the phototransistor contains a daylight blocking filter in order to reduce the sunlight radiation. The use of the blue LED in environmental conditions resulted that this device can attenuate the radiation that is generated by the sunlight at 2 V, and can detect our reference of source that was used in the experiments. It is important to mention that gain of the blue LED and phototransistor was clipped off at 5.2 V because of the power supply limitation. It is clear, that it is possible to avoid the saturation of the op amp by incrementing the distance between the OSS and the source of light. However, it was not done for reasons of standardization of the experiments. Table 9. Experiment in a laboratory with bulb source light and photodiode OP301 used as photosensor. Laboratory Test Position 90◦

OPT301 Blue LED Infrared LED LDR Phototransistor

Position 91◦

Position 92◦

σ

µ

σ

µ

σ

µ

0.215 0.138 0.305 0.264 0.325

90.397 89.83 90.4 89.763 90.079

0.196 0.156 0.307 0.352 0.334

91.6 90.778 90.623 90.8 91.222

0.169 0.206 0.285 0.313 0.406

92.682 92.009 91.399 91.646 92.441

Sunlight Conditions Position 90◦ σ OPT301 Blue LED Infrared LED LDR Phototransistor

Position 91◦

µ

σ

89.94

0.11

* 0.16

µ

σ

91.53

0.11

*

* * 0.15

Position 92◦

*

* * 89.61

0.22

µ 91.97 * *

90.72

0.19

91.78

* The OSS not detected the signal caused by source of reference.

On the other hand, the devices, such as OPT301, LDR, and infrared LED, cannot detect the reference of source, because the sunlight saturated the sensor according to their spectral responsivity, see Table 9. The best response of the photodiode OPT301 is the over spectral range (700 nm to 800 nm) according to a datasheet in the infrared region. In the region of visible light, the OPT301 has high responsivity at 650 nm 0.47 A/W. However, it cannot detect the light bulb because of interference with sunlight. This device can be limited to work in real-life environmental conditions, but we can get satisfactory results in a controlled environment, such as a laboratory. When the LDR was used and infrared for the experiment out of the laboratory the results were null. The perspective to work is controlling the angular velocity of motor due to the OSS accuracy and resolution depends on the stability of the measurement speed. Additionally, one of the main goals is to implement and develop a system that is based on microcontroller by employing digital filters that can be used for various practical applications, such as large engineering structures SHM, mobile robot navigation, remote sensing, etc. Author Contributions: Wendy Flores-Fuentes established the method in this paper in order to find the energy center; Jesús Elías Miranda-Vega wrote the paper and realized the experiments; Moisés Rivas-López analyzed the data; Oleg Sergiyenko designed the experiments with other types of photosensors; Julio C. Rodríguez-Quiñonez proposed the experimental set-up for the evaluation of the optical scanning system; Lars Lindner revised the final version of the paper. Acknowledgments: This work was supported by CONACYT and Instituto de Ingeniería de la Universidad Autónoma de Baja California. Conflicts of Interest: The authors declare no conflict of interest.

Sensors 2018, 18, 1684

15 of 15

References 1. 2. 3.

4. 5. 6. 7. 8. 9. 10.

11.

12. 13. 14. 15. 16.

17.

18.

Sohn, H.; Farrar, C.R.; Hemez, F.M.; Shunk, D.D.; Stinemates, D.W.; Nadler, B.R.; Czarnecki, J.J. A Review of Structural Health Monitoring Literature: 1996–2001; Los Alamos National Laboratory: Los Alamos, NM, USA, 2004. Schowengerdt, R.A. Remote Sensing: Models and Methods for Image Processing; Elsevier Science: New York, NY, USA, 2006. Geist, T.; Höfle, B.; Rutzinger, M.; Pfeifer, N.; Stötter, J. Laser scanning—A paradigm change in topographic data acquisition for natural hazard management. In Sustainable Natural Hazard Management in Alpine Environments; Veulliet, E., Johann, S., Weck-Hannemann, H., Eds.; Springer: Berlin/Heidelberg, Germany, 2009; pp. 309–344. Mims, F.M. The Forrest Mims Circuit Scrapbook; LLH Technology Pub.: Eagle Rock, VA, USA, 2000. Sergiyenko, O.; Rodriguez-Quiñonez, J.C. Developing and Applying Optoelectronics in Machine Vision; IGI Global: Hershey, PA, USA, 2016. Brooks, D.R.; Mims, F.M. Development of an inexpensive handheld led-based sun photometer for the globe program. J. Geophys. Res. Atmos. 2001, 106, 4733–4740. [CrossRef] Mims, F.M. Solar radiometer with light-emitting diodes as spectrally-selective detectors. Appl. Opt. 2000, 39, 6517–6518. [CrossRef] Raharijaona, T.; Mawonou, R.; Nguyen, T.; Colonnier, F.; Boyron, M.; Diperi, J.; Viollet, S. Local positioning system using flickering infrared leds. Sensors 2017, 17, 2518. [CrossRef] [PubMed] Sergiyenko, O.; Hernandez, W.; Tyrsa, V.; Cruz, L.F.D.; Starostenko, O.; Peña-Cabrera, M. Remote sensor for spatial measurements by using optical scanning. Sensors 2009, 9, 5477–5492. [CrossRef] [PubMed] Miranda-Vega, J.E.; Flores-Fuentes, W.; Sergiyenko, O.; Rivas-López, M.; Lindner, L.; Rodríguez-Quiñonez, J.C.; Hernández-Balbuena, D. Optical cyber-physical system embedded on an fpga for 3d measurement in structural health monitoring tasks. Microprocess. Microsyst. 2018, 56, 121–133. [CrossRef] Rivera-Castillo, J.; Flores-Fuentes, W.; Rivas-López, M.; Sergiyenko, O.; Gonzalez-Navarro, F.F.; Rodríguez-Quiñonez, J.C.; Hernández-Balbuena, D.; Lindner, L.; Básaca-Preciado, L.C. Experimental image and range scanner datasets fusion in shm for displacement detection. Struct. Control Health Monit. 2017, 24, e1967. [CrossRef] Shao, W.; Zhang, H.; Zhou, H. Fine particle sensor based on multi-angle light scattering and data fusion. Sensors 2017, 17, 1033. [CrossRef] [PubMed] Liu, S.C. Analog Vlsi: Circuits and Principles; MIT Press: Cambridge, MA, USA, 2002. Maini, A.K. Lasers and Optoelectronics: Fundamentals, Devices and Applications; Wiley: Hoboken, NJ, USA, 2013. Fayer, M.D. Absolutely Small Chapter 4: The Photoelectric Effect and Einstein’s Explanation; AMACOM: New York, NY, USA. Lopez, M.R.; Sergiyenko, O.Y.; Tyrsa, V.V.; Perdomo, W.H.; Cruz, L.F.D.; Balbuena, D.H.; Burtseva, L.P.; Hipolito, J.I.N. Optoelectronic method for structural health monitoring. Struct. Health Monit. 2010, 9, 105–120. [CrossRef] Rivas, M.S.; Flores, W.; Rivera, J.; Sergiyenko, O.; Hernández-Balbuena, D.; Sánchez-Bueno, A. A method and electronic device to detect the optoelectronic scanning signal energy centre. In Optoelectronics—Advanced Materials and Devices; Pyshkin, S.L., Ballato, J.M., Eds.; InTech: Rijeka, Croatia, 2013; Chapter 16. De Oliveira, M.; Araujo, N.; da Silva, R.; da Silva, T.; Epaarachchi, J. Use of savitzky–golay filter for performances improvement of SHM systems based on neural networks and distributed PZT sensors. Sensors 2018, 18, 152. [CrossRef] [PubMed] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).