Design and Development of a Low-Cost Automated All-Terrain Intelligent Robotic Vehicle for Detection to Study Its Faults and Vulnerabilities from † SWOT Perspective

proceedings

Proceedings Design and Development of a Low-Cost Automated All-Terrain Intelligent Robotic Vehicle for Detection to Study Its Faults and Vulnerabilities from † SWOT Perspective

Florin Covaciu 1, Persida Bec 2 and Doru-Laurean Băldean 1,*

1 Design Engineering and Robotics Department, Faculty of Machines Building, Technical University of Cluj-Napoca, 103-105 Muncii, 400641 Cluj-Napoca, Romania; ﬂ[email protected] 2 Ethics of Vulnerabilities Group, Faculty of Philosophy, Babes-Bolyai University, Kogălniceanu 1, 400084 Cluj-Napoca, Romania; [email protected] * Correspondence: [email protected] or [email protected]; Tel.: +40-26-420-2790 Presented at the 14th International Conference on Interdisciplinarity in Engineering—INTER-ENG 2020, † Târgu Mures, , Romania, 8–9 October 2020.

Published: 18 December 2020

Abstract: This scientiﬁc article highlights the development of an all-terrain intelligent robotic vehicle (ATIRV) platform, its overall structure, virtual modeling and simulation programs (such as Unity 5), electronic controls, and corresponding modules. This paper also shows the physical structure developed for the all-terrain intelligent vehicle and some of its components, which allow other modern robotic vehicles to operate autonomously and automatically in uncharted and dangerous environments. The practical part of the project has been developed by using state-of-the-art features like tele-metrics, tele-operation, drive assist modules, and autonomous navigation using the physical model of the all-terrain intelligent vehicle as structural model in the research experiment and for validating the initial hypothesis (of automatic ATIRV). Other scientiﬁc contributions consist in the determination of operational (kinematic and dynamic) parameters. Their comparison with reference values guides the paper’s discussions and conclusions.

Keywords: automation; intelligent systems; design; manufacturing; robotic vehicles; vulnerability

1. Introduction Design and development, supported by computational power, are already applied in the optimization of robotic systems [1]. Hazardous environments involve the usage of robotic applications for the detection of combat-mines, explosives, and dangerous materials to save lives and to protect the health of civilians and personnel engaged in decontamination. Applied solutions concerning robotic systems for quasi-autonomous inspections [2], dangerous material detection/measurement [3], harshness assessment [4], personal safety [5], automated controlled vehicles [6], and Strengths, Weaknesses, Opportunities, and Threats (SWOT) analysis of the development processes [7] have been partially approached, and applied in laboratory research. They also have been field-tested in given conditions to some extent [8,9]. Dangerous fields, such as postwar areas, roads, agricultural terrains, and accident sites with fatal casualties, are seen to be the beneficiaries of remotely controlled robotic applications. If combat-mines or sharp or cutting metals exist in the surface soil, then that field has a high degree of risk potential in other civil applications (injury may occur and fatalities and bodily integrity threats are possible). Analyzing applied data from cross-sectional testing of operational

Proceedings 2020, 63, 38; doi:10.3390/proceedings2020063038 www.mdpi.com/journal/proceedings Proceedings 2020, 63, 38 2 of 11 parameters is a request in the fields of vehicle use, terrain exploration, and other utility equipment [10]. Studies of smart control of the operational systems installed on the utility vehicles [11], and the optimal planning of displacement, in the case of mobile robots, using mathematical algorithms [12] as well as Arduino and other platforms [13–15], were applied for specific propulsion and robot-in-the-loop coordination [16]. Resource protection and security are two of the main objectives for these kinds of developments. Enhanced power sources are a key factor [17] for autonomous robotic vehicles [18] in hazardous scenarios [19]. The primary objectives of the present paper include designing an all-terrain intelligent robotic vehicle (ATIRV); developing of a low-cost model for metal detection; studying the faults, challenges, and vulnerabilities of the laboratory model; extending the capabilities of the laboratory work to the real-size vehicle; and presenting the first test results. The ATIRV design process, development, and testing sequences cover aspects regarding multiple operating modes (such as accelerate, drive, brake, decelerate, and detect), as well as engineering innovation solutions for reaching the primary objective of research. This research addresses problems and aspects related to designing, programming, sensor installation, remote digital management, vehicle electronic control, actuator structures, and multiple practical testing (both in laboratory conditions with a low-cost mini-model and in the field scenario with a real-size all-terrain vehicle). These scenarios have been made to verify the physical platform and the predefined algorithms for automated driving. Challenges, faults, and vulnerabilities of the robotic models were also closely monitored during the design and development process. The intelligent components installed in this application consist in fuzzy regulators which are used in the vehicle’s control part. First, a virtual reality model for an experimental mobile robot with field exploration and detection capabilities was created. Second, the mini-model for laboratory testing was realized. The third phase presented the real-size vehicle dynamic field testing in which the programming and digital control were applied to an actual All-Terrain Vehicle (ATV) for data recording and scientific study of the problem. The main tasks of ATIRVs are field exploration and the detection of metal objects (like combat-mines).

2. Materials and Designing Method The applied scientific work is supported by an adaptable design of an automated robotic all-terrain vehicle that is designed to detect specific objects in broken or hazardous terrains. Considering these aspects is a quite often debated and defined problem in robotics and mechatronics, especially when high mobility or autonomy is required. Significant demands for this kind of robotic applications are also taken into consideration in domains regarding victim assistance, complex fires, atomic hazards, and space exploration. During the sequence of ATIRV designing and project development, engineering-specialized programs (such as Unity 5, MATLAB/Simulink, AutoCAD, and SolidWorks) were used to study both the one-degree model and the all-terrain-wheeled robot with mathematical equation systems. A series of calculations and programming were done to determine and control the operational dynamic parameters related to robotic vehicles by using a digital platform. A simple diagram with the connections of the components is shown in Figure1. Virtual reality (VR) is the cutting edge of technology development. Using VR tools facilitates the experimentation of a new reality (that does not exist in material world). It is used in important applications (such as the ones mentioned and proposed in the present paper). Part of the design for ATIRVs is supported by the Siemens NX program. In the first place, the robot’s platform was generated. This was provided with “home position” for the robotic arm, headlights, energy source, and antenna. After the virtual creation of the robot’s platform, the wheel train was designed. Then the components were assembled and a format conversion was made to facilitate implementation in the virtual reality. After the 3D construction and virtual modeling of the mobile robot, it was imported into the Unity 5 environment, which supports specific formats and extensions such as “3ds”, “dae”, “dxf”, “fbx”, “obj”, and “skp”. The next step was to assign material properties to the components of the mobile robot in the Keyshot program. The main challenge was the size of the resulting file, which was Proceedings 2020, 63, 38 3 of 11

Proceedings 2020, 4, x FOR PEER REVIEW 3 of 11 considerably large. Due to the problems in using such a volume-demanding format, the alternative 3D modeling, animation, rendering, and graphical composition in motion. All the components must 3ds MaxProceedings program 2020, became4, x FOR PEER appealing. REVIEW This software solution allowed complex 3D modeling, animation,3 of 11 be verified in 3ds Max to have a correct load. Other faults and challenges were found in the fact that rendering, and graphical composition in motion. All the components must be verified in 3ds Max to the 3ds3D Max modeling, program animation, did not rendering, keep the and restraints graphical fr omcomposition Siemens in NX. motion. The positionsAll the components of the axles must were, have a correct load. Other faults and challenges were found in the fact that the 3ds Max program did at thisbe point, verified misplaced. in 3ds Max Beside to have all a the correct designing load. Other process, faults one and specific challenges innovation were found was in the factsolution that for not keep the restraints from Siemens NX. The positions of the axles were, at this point, misplaced. aligningthe all3ds the Max axles program in the did correct not keep positions. the restraints We asfromsigned Siemens coordinates NX. The positionsto each part of the of axles axis. were, Then we Beside all the designing process, one specific innovation was the solution for aligning all the axles in replacedat this all point, the axles misplaced. of components Beside all the according designing to pr theocess, topological one specific data innovation from Siemens was the solution NX, obtaining for the correct positions. We assigned coordinates to each part of axis. Then we replaced all the axles of in thisaligning manner all the the correct axles in positions the correct of positions. the corresponding We assigned components. coordinates toThis each was part done of axis. by selectingThen we the componentsreplaced according all the axles to of the components topological according data from to the Siemens topological NX, obtainingdata from inSiemens this manner NX, obtaining the correct specific part from 3ds Max, entering the pivot menu, and changing the axis position. After assigning positionsin this of manner the corresponding the correct positions components. of the corresponding This was done components. by selecting This the was specific done partby selecting from 3ds the Max, physical material to each component, the importation process needed to be realized into the Unity 5 enteringspecific the part pivot from menu, 3ds Max, and changingentering the the pivot axis me position.nu, and changing After assigning the axis position. physical After material assigning to each application. component,physical the material importation to each component, process needed the importatio to be realizedn process into needed the Unity to be 5 realized application. into the Unity 5 application.

Figure 1. Simplified structure of the physical components (a) and the control algorithm (b) for all- Figure 1.1. SimpliﬁedSimplifiedstructure structure of of the the physical physical components components (a) and(a) and the controlthe control algorithm algorithm (b) for (b all-terrain) for all- terrain intelligent robotic vehicles (ATIRVs). terrainintelligent intelligent robotic robotic vehicles vehicles (ATIRVs). (ATIRVs).

An importantAn important step step in thein the design design and and virtual virtual developmentdevelopment of of the the ATIRV ATIRV consisted consisted in material in material Anassignation important and step rendering in the from design wireframe and virtual and coating development parameter of definition, the ATIRV as shown consisted in Figure in material 2. assignation and rendering from wireframe andand coatingcoating parameterparameter deﬁnition,definition, asas shownshown inin FigureFigure2 .2.

(a) (b)

Figure 2. Virtual modeling with material map browser: (a) assigning materials to virtual components; (b) rendering of robotic vehicle(a) from wireframe after the components were “materialized”.(b) Figure 2. VirtualVirtual modeling with material map browser: ( (aa)) assigning assigning materials materials to to virtual virtual components; components; The Unity 5 environment is used for 3D games and VR simulators, allowing 3 programming (b) rendering ofof roboticrobotic vehiclevehicle fromfrom wireframewireframe afterafter thethe componentscomponents werewere “materialized”.“materialized”. codes (C#, JavaScript, and Boo) and exporting capabilities for multiple operating systems. Prior to ATIRV import into the Unity 5 application, a textured environment was generated using the Gaia The Unity 5 environment is used for 3D games and VR simulators, allowing 3 programming codes Theprogram, Unity which 5 environment allows the creation is used of for vegetation, 3D game was ter,and and VR rocks simulators, (of different allowing sizes and 3 shapes),programming as (C#,codes JavaScript,shown (C#, JavaScript, in Figure and Boo)3. and and Boo) exporting and exporting capabilities capabilities for multiple for multiple operating operating systems. systems. Prior to Prior ATIRV to ATIRVimport intoimport the into Unity the 5 application,Unity 5 application, a textured a textured environment environment was generated was generated using the using Gaia program,the Gaia whichprogram, allows which the allows creation the ofcreation vegetation, of vegetation, water, and wa rockster, and (of rocks diﬀerent (of different sizes and sizes shapes), and shapes), as shown as shownin Figure in3 Figure. 3.

Proceedings 2020, 63, 38 4 of 11 Proceedings 2020, 4, x FOR PEER REVIEW 4 of 11 Proceedings 2020, 4, x FOR PEER REVIEW 4 of 11

Figure 3. Virtual modeling of the sustaining environment in the Gaia program for implementation Figure 3. Virtual modelingmodeling of of the the sustaining sustaining environment environment in thein the Gaia Gaia program program for implementationfor implementation into into Unity 5. intoUnity Unity 5. 5.

The UnityUnity 55 applicationapplication may may be be generated generated for for di ﬀdierentfferent types types of operatingof operating systems systems (Android, (Android, iOS, The Unity 5 application may be generated for different types of operating systems (Android, iOS,Linux, Linux, Mac, Mac, PC, PS4,PC, PS4, Xbox, Xbox, etc.), etc.), as shown as shown in Figure in Figure4. 4. iOS, Linux, Mac, PC, PS4, Xbox, etc.), as shown in Figure 4.

Figure 4. Importing the Gaia environment in Unity 5 and the Operating System (OS) selection. Figure 4. ImportingImporting the the Gaia Gaia environment environment in Unity 5 and the Operating System (OS) selection. With the support of the developed virtual model, a laboratory mobile mini-robot was realized With the support of the developeddeveloped virtual model, a laboratory mobile mini-robot was realized to test similar features, such as detection and exploration, for the ATIRV. The software application to test similar features, such as detection and and exploration, exploration, for for the the ATIRV. ATIRV. TheThe softwaresoftware applicationapplication facilitated the virtual modeling of different robotic mechanisms. Robotic design allowed an improved facilitated the virtual modeling of different different roboticrobotic mechanisms. Robotic design allowed an improved control of the automation suitable processes. The design and development of the lab model was control of the automationautomation suitable processes.processes. The The design design and and development of of the lab model was strongly facilitated by the virtual construction of the automated vehicle. The virtual model may be strongly facilitated by the virtual construction of the automatedautomated vehicle. The virtual model may be designed and simulated on the same computer station, but the physical model has two distinctive designed and simulated on the same computer station, but the physical model has two distinctive control units, one being the server (at the programming level) and the other being the client Espressif control units, one being the server (at the programmi programmingng level) and the other being the client Espressif Systems Part (ESP 32 micro-controller). Systems Part (ESP 32 micro-controller). SystemsThe Partelectronic (ESP 32control micro-controller). unit acquires all the signals and information from the transferring module The electronic control unit acquires all the signals and information from the transferring module once Thethe transfer electronic procedure control unit is complete acquires and all the will signals send the and signals information toward from the thepowertrain transferring assembly module to once the transfer procedure is complete and will send the signals toward the powertrain assembly to driveonce thethe transferATIRV to procedure the new isdefined complete location. and will In rela sendtion the to signals the physical toward road the powertraintrack, the method assembly of drive the ATIRV to the new defined location. In relation to the physical road track, the method of detectingto drive the and ATIRV following to the the new magnetic defined material location. lane In relationis very efficient to the physical and reliable road in track, this type the method of robotic of detecting and following the magnetic material lane is very efficient and reliable in this type of robotic applicationdetecting and for following reaching thethe magneticexpected materialoutcome lane and ispr veryoper e results.fficient andThe reliableoptimal in driving this type track of roboticlane is application for reaching the expected outcome and proper results. The optimal driving track lane is realizedapplication when for taking reaching into the consideration expected outcome all the aspects and proper and vulnerabilities results. The optimal that may driving occur. track The ATIRV lane is realized when taking into consideration all the aspects and vulnerabilities that may occur. The ATIRV wasrealized a battery-propelled when taking into considerationelectric vehicle all thewhich aspects could and charge vulnerabilities with electricity that may automatically. occur. The ATIRV To was a battery-propelled electric vehicle which could charge with electricity automatically. To wasimprove a battery-propelled the ATIRV operational electric vehicle performance, which couldinductive charge power with transfer electricity methods automatically. were applied. To improve improve the ATIRV operational performance, inductive power transfer methods were applied. the ATIRVMaterial operational components performance, used by inductivethe present power work transfer consist methods in electronic were applied. software programs, Material components used by the present work consist in electronic software programs, commandMaterial units, components and physical used actuators by the for presentcontrolling work the consistautomated in electronicvehicle (specific software data programs,regarding command units, and physical actuators for controlling the automated vehicle (specific data regarding technicalcommand features units, and are physicaloffered in actuators Table 1). for controlling the automated vehicle (specific data regarding technical features are oofferedffered inin TableTable1 ).1). Table 1. Specific data regarding the robotic vehicle materials used for research. Table 1. SpecificSpecific data regarding the robotic vehicle materials used for research. Programming Level Command Level Actuator Level Programming Level Command Level Actuator Level OperatingProgramming system Level Metal Command detector Level Actuator Level Feetech Operating system Metal detector Feetech ArduinoOperating system GPS module; Metal TTLinker detector Board Feetech SCS 15 Arduino GPS module; TTLinker Board SCS 15 C++ Arduino ESPGPS 32; module; Lipo Accumulator TTLinker Board SCS Servomotor 15 C++ C++ ESPESP 32; 32; Lipo Lipo Accumulator Accumulator Servomotor Servomotor

ProceedingsProceedings 20202020,, 44,, xx FORFOR PEERPEER REVIEWREVIEW 55 ofof 1111

3.3. DevelopmentDevelopment andand ResultsResults

ProceedingsProximityProximity2020, 63 , sensingsensing 38 transducerstransducers werewere installedinstalled inin thethe roboticrobotic ATIRVATIRV drivewaydriveway inin orderorder toto gaingain5 of 11 informationinformation aboutabout thethe vehicle’svehicle’s surroundingsurrounding obstacleobstacles,s, kinetickinetic actions,actions, distances,distances, andand emptyempty paths,paths, thethe kindskinds ofof datadata whichwhich practicallypractically imposeimpose thethe startingstarting andand stoppingstopping sequences.sequences. PhotometricPhotometric 3.transducerstransducers Development werewere and setset upup Results toto gaingain informationinformation concerniconcerningng obstaclesobstacles oror objectsobjects onon thethe track.track. AA predefinedpredefined objectobjectProximity detectiondetection sensing modulemodule transducers waswas alsoalso provided,provided, were installed andand aa physicalphysical in the robotic transfertransfer ATIRV systemsystem driveway facilitatedfacilitated in detectiondetection order to andand gain deactivation of hazardous materials with a package of special mechatronic equipment, in which all informationdeactivation about of hazardous the vehicle’s materials surrounding with a package obstacles, of special kinetic mechatronic actions, distances, equipment, and emptyin which paths, all electric and electronic connections were interdependent. theelectric kinds and of data electronic which practicallyconnections impose were interdependent. the starting and stopping sequences. Photometric transducers The first phase of development consisted in design, calculus of dimensions, and virtual were setThe up first to gainphase information of development concerning consisted obstacles in design, or objects calculus on the of track. dimensions, A predefined and virtual object development using Computer Aided Design (CAD) tools, as shown in Figure 5. detectiondevelopment module using was Computer also provided, Aided Design and (CAD) a physical tools, transferas shown system in Figure facilitated 5. detection and TheThe secondsecond importantimportant phasephase consistedconsisted inin materialmaterial acquirementacquirement andand wirewire connections.connections. ElectronicElectronic deactivation of hazardous materials with a package of special mechatronic equipment, in which circuitscircuits andand sensorsensor connectionsconnections werewere mademade toto developdevelop thethe detectiondetection system,system, whichwhich waswas providedprovided all electric and electronic connections were interdependent. withwith LEDLED signalingsignaling lamplamp 1,1, forfor objectobject presence,presence, andand signalsignal amplifieramplifier 2,2, asas shownshown inin FigureFigure 6.6. The first phase of development consisted in design, calculus of dimensions, and virtual TheThe thirdthird phasephase ofof thethe designdesign andand developmentdevelopment processprocess consistedconsisted inin vehiclevehicle assemblyassembly development using Computer Aided Design (CAD) tools, as shown in Figure5. (servomotor(servomotor 1,1, TTLinkerTTLinker boardboard 2,2, conversionconversion boardboard 3)3) andand detectiondetection testing.testing. ForFor thethe powerpower supplysupply ofof The second important phase consisted in material acquirement and wire connections. Electronic thethe robot,robot, aa LipoLipo 7.47.4 V/1300V/1300 mAmA accumulatoaccumulatorr 44 waswas provided.provided. TheThe hardwarehardware structurestructure waswas circuits and sensor connections were made to develop the detection system, which was provided with manufacturedmanufactured andand assembledassembled forfor ATIRVATIRV completecomplete developmentdevelopment andand forfor powerpower supplysupply testing,testing, asas LED signaling lamp 1, for object presence, and signal amplifier 2, as shown in Figure6. shownshown inin FigureFigure 7.7.

((aa)) ((bb))

FigureFigureFigure 5. 5.5. Virtual Virtual modelmodel ofof thethe platform platform for for minimini roboticrobotic vehicle:vehicle: ((a (aa)) ) 2D2D 2D drawingdrawing drawing ofof of thethe the robotrobot robot platform;platform; platform; ((b(bb))) 3D 3D3D model modelmodel of ofof ATIRVATIRVATIRV main mainmain platformplatformplatform basedbased on on a a commercial commercialcommercial kitkit ofof 4-wheeled4-wheeled 4-wheeled remoteremote remote controlledcontrolled controlled car.car. car.

((aa)) ((bb))

FigureFigureFigure 6. 6.6. ElectronicElectronic connectionconnection betweenbetween the the GPS GPS modulemodule anan anddd thethe microcontroller,microcontroller, microcontroller, asas wellwell well asas as additionsadditions additions tototo the thethe metalmetalmetal sensor: sensor:sensor: ((a (aa)) )wiringwiring wiring ofof of thethe the GPSGPS GPS structurestructure structure withwith with thethe themicrocontrollermicrocontroller microcontroller board;board; board; ((bb)) detector.detector. (b) detector.

The third phase of the design and development process consisted in vehicle assembly (servomotor 1, TTLinker board 2, conversion board 3) and detection testing. For the power supply of the robot, a Lipo 7.4 V/1300 mA accumulator 4 was provided. The hardware structure was manufactured and assembled for ATIRV complete development and for power supply testing, as shown in Figure7. A switch was added to turn the robot on whenever required and to turn it oﬀ when the detection task was completed, for energy conservation in the accumulator. One fault of the GPS module, which was outlined during the tests, consisted in the duration of about 15 min it took to establish satellite connection outdoor. Indoor, the connecting process became even harder, and it could take up to one hour. This made robot testing a challenge because we had to make the connection ﬁrst, prior to running the detection program. In light of the above-mentioned problem, the robot’s antenna was Proceedings 2020, 4, x FOR PEER REVIEW 6 of 11

(a) (b)

Figure 7. Platform development of an automated all-terrain digitally controlled vehicle for detection: (a) basic platform assembly; (b) testing of energy supply system and digital command.

A switch was added to turn the robot on whenever required and to turn it off when the detection task was completed, for energy conservation in the accumulator. One fault of the GPS module, which Proceedingswas outlined2020, 63 during, 38 the tests, consisted in the duration of about 15 min it took to establish satellite6 of 11 connection outdoor. Indoor, the connecting process became even harder, and it could take up to one hour. This made robot testing a challenge because we had to make the connection first, prior to installedrunning above the detection the plastic program. case of In the light electronic of the components,above-mentioned to facilitate problem, satellite the robot’s connection, antenna as was shown ininstalledProceedings Figure8 . above2020, 4, xthe FOR plastic PEER REVIEW case of the electronic components, to facilitate satellite connection,6 of as 11 shown in Figure 8. The hardware platform we received with the Arduino kit and initially used had some major issues, a fact which contributed to the personalized design of the mobile robot. Thus, additional pockets were created in the platform to attach the servomotors for the front D65X28 wheels. Another adjustment we had to make was the creation of a section where we could attach the robotic detector tool, since the standard design version of the kit did not include such equipment. The final challenge was the acrylic plastic material of structural platform from the standard assembly because it would collapse under the weight and stress of the detector equipment. The power consumption was calculated using Equation (1). Using a software interface, the user can define specific IDs to each servomotor. The default ID is 0. To change the ID with( thisa) program the “unlock” checkbox on top must (beb) clicked, the ID value in the FigurecorrespondingFigure 7. 7.Platform Platform field development development written, “set” ofof anan pressed, automated an all-terraind all-terrain then “WRITE” digitally digitally controlledhit. controlled Additionally, vehicle vehicle for the for detection: detection:servo type may(a )be( basica) setbasic platformto platform servo-mode assembly; assembly; or to ( b( b)wheel) testing testing mode, ofof energyenergy and supplysupplymany system other pre-testing and and digital digital command.specifications command. may be set, by using script-fields and selection boxes for data input, as shown in Figure 9. A switch was added to turn the robot on whenever required and to turn it off when the detection task was completed, for energy conservation in the accumulator. One fault of the GPS module, which was outlined during the tests, consisted in the duration of about 15 min it took to establish satellite connection outdoor. Indoor, the connecting process became even harder, and it could take up to one hour. This made robot testing a challenge because we had to make the connection first, prior to running the detection program. In light of the above-mentioned problem, the robot’s antenna was installed above the plastic case of the electronic components, to facilitate satellite connection, as shown in Figure 8. The hardware platform we received with the Arduino kit and initially used had some major issues, a fact which contributed(a) to the personalized design of the mobile(b) robot. Thus, additional pocketsFigureFigure were 8. 8.All-terrain All-terrain created in intelligentintelligent the platform robotic to vehicle vehicleattach modelthe model servomotors for for detection detection for ready ready the for front for testing testing D65X28 and and optimization: wheels. optimization: Another adjustment(a()a virtual) virtual we model; model; had (to b(b) )make laboratory laboratory was the prototypeprototype creation model model of a for forsection testing testing where and and development. development.we could attach the robotic detector tool, since the standard design version of the kit did not include such equipment. The final challenge wasThe the hardware acrylic plastic platform material we of received structural with platform the Arduino from the kit standard and initially assembly used because had some it would major issues,collapse a fact under which the contributedweight and tostress the personalizedof the detector design equipment. of the mobileThe power robot. consumption Thus, additional was pocketscalculated were using created Equation in the platform(1). to attach the servomotors for the front D65X28 wheels. Another adjustmentUsing we a software had to make interface, was the the creation user can of define a section specific where IDs we to couldeach servomotor. attach the robotic The default detector ID is tool, since0. To the change standard the ID design with versionthis program of the the kit “unlock” did not includecheckbox such on top equipment. must be clicked, The final the challenge ID value in was thethe acrylic corresponding plastic material field written, of structural “set” platformpressed, fromand then the standard“WRITE” assembly hit. Additionally, because itthe would servo collapse type undermay thebe set weight to servo-mode and stress or of theto wheel detector mode, equipment. and many The other power pre-testing consumption specifications was calculated may be set, using Equationby using (1). script-fields and selection boxes for data input, as shown in Figure 9. Using a software interface, the user can define specific IDs to each servomotor. The default ID is 0. To change the ID with this program the “unlock” checkbox on top must be clicked, the ID value in the corresponding field written, “set” pressed, and then “WRITE” hit. Additionally, the servo type may be set to servo-mode or to wheel mode, and many other pre-testing specifications may be set, by using script-fields and selection boxes for data input, as shown in Figure9. Proceedings 2020, 4, x FOR PEER REVIEW 7 of 11

(a) (b)

Figure 8. All-terrain intelligent robotic vehicle model for detection ready for testing and optimization: (a) virtual model; (b) laboratory prototype model for testing and development.

(a) (b)

FigureFigure 9. Feetech9. Feetech FD-v1.8 FD-v1.8 software software application:application: (a ()a )defining deﬁning the the ID IDof servo of servo motor; motor; (b) menu (b) menu for for kinematickinematic and and dynamic dynamic parameters parameters deﬁnition. definition.

Microsoft Visual Studio 2017 was used for the design and development of the graphical user interface with map display system and client-server communication, as shown in Figure 10. The server in this case is on the PC station inside the control room and the client is on the mobile robot that is remotely driven through the smart application.

(a) (b)

Figure 10. User-adapted graphic interface and the communication tool between the client and server: (a) coordinates definition; (b) data saving tool in the user interface and map location monitoring.

For experimental testing of the proposed idea was considered a research-prepared ATIRV, as shown in Figure 11.

Figure 11. Simplified schematic (a) of the experimental proposed ATIRV (b) consisting of the following main components: (1) front sensor; (2) sensor support; (3) conversion board and wire assembly box; (4) experimental configured ATIRV; (5) GPS module safety box; (6) accumulator safety box.

The hardware assembly we received with the experimental ATV kit and first used in testing had multiple major issues, a fact which contributed to the specialized design of the ATIRV. Thus, additional components were created and placed in the front of the vehicle to attach the metal detecting sensor, the wiring box, the GPS protection case, and the accumulator safety cover. Additionally, the ABS (Antilock Braking System) needed to be programmed as BAS (Braking Assist System) to improve stopping control of the wheels. Another contribution we had to make consisted of an adjustment where the metal detector sensor was attached, since the standard design version of the kit did not include such equipment. After solving the practical challenges, tests were performed. During the experimental testing, kinematic data were obtained by using GPS data and acceleration sensor, which are significant data of ATIRV operation.

ProceedingsProceedings 2020 2020, 4, ,4 x, xFOR FOR PEER PEER REVIEW REVIEW 7 7of of 11 11

(a()a ) (b(b) )

ProceedingsFigureFigure2020 9. , 9.63 Feetech ,Feetech 38 FD-v1.8 FD-v1.8 software software application: application: ( a()a )defining defining the the ID ID of of servo servo motor; motor; ( b(b) )menu menu for for 7 of 11 kinematickinematic and and dynamic dynamic parameters parameters definition. definition.

MicrosoftMicrosoftMicrosoft VisualVisual Visual StudioStudio Studio 2017 2017 was waswas used usedused for for the the design design and and and development development development of of ofthe the the graphical graphical graphical user user user interfaceinterfaceinterface with with with map map map displaydisplay display system system and and and client-s client-serverclient-servererver communication, communication, as as asshown shown shown in in inFigure Figure Figure 10. 10. 10 . TheTheThe server server server inin in thisthis this case case is is on on the thethe PC PCPC station stationstation inside inside the the the control control control room room room and and and the the the client client client is is on ison the on the themobile mobile mobile robotrobotrobot that that that is is is remotely remotely remotely drivendriven driven through through the the the smart smartsmart application. application.application.

(a()a ) (b(b) )

FigureFigureFigure 10. 10. 10. User-adaptedUser-adapted User-adapted graphic graphic interface interfaceinterface and and the the communi communi communicationcationcation tool tool tool between between between the the theclient client client and and andserver: server: server: (a()a()a coordinates )coordinates coordinates deﬁnition;definition; definition; ((b (b) ) data data saving saving saving tool tooltool in inin the thethe user user in in interfaceterfaceterface and and and map map map location location location monitoring. monitoring. monitoring.

ForForFor experimental experimental experimental testing testing testing of of theof the the proposed proposed proposed idea idea idea was was was considered considered considered a research-prepared a a research-prepared research-prepared ATIRV,as ATIRV, ATIRV, shown as as inshownshown Figure in in11 Figure Figure. 11. 11.

Figure 11. Simplified schematic (a) of the experimental proposed ATIRV (b) consisting of the FigureFigure 11. 11.Simplified Simplified schematic schematic (a )(a of) theof the experimental experimental proposed proposed ATIRV ATIRV (b) consisting (b) consisting of the of following the following main components: (1) front sensor; (2) sensor support; (3) conversion board and wire mainfollowing components: main components: (1) front sensor; (1) front (2) sensorsensor; support;(2) sensor (3) support; conversion (3) conversion board and board wire assembly and wire box; assembly box; (4) experimental configured ATIRV; (5) GPS module safety box; (6) accumulator safety box. (4)assembly experimental box; (4) configuredexperimental ATIRV; configured (5) GPS ATIRV; module (5) GPS safety module box; safety (6) accumulator box; (6) accumulator safety box. safety box. The hardware assembly we received with the experimental ATV kit and first used in testing had TheThe hardware hardware assemblyassembly we received received with with the the experimental experimental ATV ATV kit kit and and first first used used in testing in testing had had multiple major issues, a fact which contributed to the specialized design of the ATIRV. Thus, multiplemultiple major major issues, issues, a facta fact which which contributed contributed to the to specializedthe specialized design design of the of ATIRV. the ATIRV. Thus, additionalThus, additionaladditional componentscomponents werewere createdcreated andand placedplaced inin thethe front front ofof thethe vehiclevehicle to to attachattach thethe metalmetal components were created and placed in the front of the vehicle to attach the metal detecting sensor, detectingdetecting sensor,sensor, thethe wiringwiring box,box, thethe GPSGPS protecprotectiontion case,case, andand thethe accumulatoraccumulator safetysafety cover.cover. the wiring box, the GPS protection case, and the accumulator safety cover. Additionally, the ABS Additionally,Additionally, the the ABS ABS (Antilock (Antilock Braking Braking System) System) ne needededed to to be be programmed programmed as as BAS BAS (Braking (Braking Assist Assist (Antilock Braking System) needed to be programmed as BAS (Braking Assist System) to improve System)System) to to improve improve stopping stopping control control of of the the wheels. wheels. Another Another contribution contribution we we had had to to make make consisted consisted stopping control of the wheels. Another contribution we had to make consisted of an adjustment ofof an an adjustment adjustment where where the the metal metal detector detector sensor sensor was was attached, attached, since since the the standard standard design design version version of of wherethethe kit kit the did did metal not not include include detector such such sensor equipment. equipment. was attached, After After solv solv sinceinging the the practical practical standard challenges, challenges, design version tests tests were were of theperformed. performed. kit did not includeDuringDuring such equipment. thethe experimentalexperimental After solving testing,testing, the kinematickinematic practical data challenges,data werewere obtainedobtained tests were byby performed. usingusing GPSGPS datadata andand accelerationaccelerationDuring the sensor, sensor, experimental which which are are testing, sign significantificant kinematic data data of of data ATIRV ATIRV were operation. operation. obtained by using GPS data and acceleration sensor, which are significant data of ATIRV operation. Longitudinal accelerations and speed variations on horizontal terrain were graphically represented for two distinctive testing scenarios ((a) without any added load; (b) with a load of 1000 N), as shown in Figure 12. The actual values obtained in the process of experimental testing of the practically prepared ATIRV, without any added load, are offered in Table2. The real values gained in the process of experimental testing of the practically configured ATIRV, with 1000 N added load, are offered in Table3. The required power was defined using the following mathematical model to gain values:

. . P = Fi v k, (1) Proceedings 2020, 4, x FOR PEER REVIEW 8 of 11

Longitudinal accelerations and speed variations on horizontal terrain were graphically represented for two distinctive testing scenarios ((a) without any added load; (b) with a load of 1000 N), as shown in Figure 12. The actual values obtained in the process of experimental testing of the practically prepared ATIRV, without any added load, are offered in Table 2. The real values gained in the process of experimental testing of the practically configured ATIRV, with 1000 N added load, are offered in Table 3.

ProceedingsThe required2020, 63, 38power was defined using the following mathematical model to gain values:8 of 11

P = Fi.v.k, (1) wherewhere P is P ispower power consumption; consumption; Fi Fisi isinertial inertial force, force, in inNewtons; Newtons; v is v isspeed; speed; and and k kis iscorrection correction factor factor (k = 1.5).(k = 1.5).

(a) (b)

FigureFigure 12. 12. GraphicsGraphics for for longitudinal longitudinal accelerations accelerations and and speeds forfor thethe ATIRVATIRV in in practical practical testing testing conditions:conditions: (a ()a without) without any any attached attached load load on on the the vehicle; ((bb)) withwith 1000 1000 N N load load attached attached on on the the vehicle. vehicle.

TableTable 2. 2.RealReal values values gained gained in in practical practical testing testing of the experimentalexperimental ATIRV ATIRV without without added added load. load.

Road’sRoad’s Conﬁguration, Configuration, Surface,Surface, Speed,Speed, Max.Max. Deceleration, Deceleration, Max.Max. Acceleration, Acceleration, 2 2 [-][-] [-][-] [km/h][km/h] [m/s[m2/]s ] [m/s[m/s2]] 1010 11.42 11.42 10.91 10.91 IrregularIrregular HorizontalHorizontal terrain terrain 2020 10.21 10.21 09.48 09.48 surfacesurface 3030 08.82 08.82 08.55 08.55

TableTable 3. 3.ActualActual values values recorded recorded in in practical practical tests of thethe experimentalexperimental ATIRV ATIRV with with added added load. load.

Road’sRoad’s Conﬁguration, Configuration, Surface,Surface, Speed,Speed, Max.Max. Deceleration, Deceleration, Max.Max. Acceleration, Acceleration, [-][-] [-][-] [km/h][km/h] [m/s[m2/]s 2] [m/s[m/s2]] 1010 12.22 12.22 09.92 09.92 IrregularIrregular HorizontalHorizontal terrain terrain 2020 11.19 11.19 08.47 08.47 surfacesurface 3030 09.52 09.52 07.91 07.91

4. 4.Discussions Discussions and and Conclusions Conclusions OneOne of of the the significant signiﬁcant innovative innovative contributions contributions of thethe presentpresent research research consists consists in in the the physical physical developeddeveloped vehicle, vehicle, which isis highly highly adaptable adaptable and and has ahas great a great degree degree of ﬂexibility, of flexibility, thus making thus it suitablemaking it suitableto be implementedto be implemented in multiple in roboticmultiple applications robotic applications and automotive and architectures.automotive architectures. The connectors, The connectors,communication communication pathways, technicalpathways, vulnerabilities, technical vulnerabilities, and safety aspects and weresafety also aspects considered were and also consideredevaluated. and This evaluated. work presented This work the design presented and results the design of control and protocolresults of for control an all-terrain protocol intelligent for an all- terrainvehicle intelligent robot which vehicle may berobot applied which in amay multitude be applied of scenarios, in a multitude such as surveillance, of scenarios, hazardous such as surveillance,transports, andhazardous metal detection, transports, in bothand metal muddy dete andction, snowy in surfaces.both muddy The ATIRVand snowy was designedsurfaces. toThe be electrically powered from a rechargeable battery package (to be used with a solar photo-voltaic

panel charger) and most of the implemented actuators were electric motors. All the latter components were electronically controlled by integrated microprocessors monitored by the vehicle’s control unit. The ATIRV was equipped with eight high-resolution digital camera sensors, 3D scanners with laser technology, a special Inertial Control Unit (ICU), a digital compass, a GPS unit, a radio transceiver kit, and long-range wireless communication capability. The primary model of the ATIRV focused on Proceedings 2020, 63, 38 9 of 11

first-level objectives such as specific detection, 3D all-terrain surveillance and mapping (on-road and off-road), basic driving styles, basic starting and stopping styles, avoiding obstacles like large rocks or stone blocks, and driving over off-road surfaces like highly fractured ground and hilly terrains. The detection tool performed specific maneuvers such as precise locating, searching, removing, or neutralizing if there were applied additional components. In relation to other research, such as on semi-autonomous robots [2], this was intended to perform fully automated, even if there was a limit regarding the complete automation of the experimental kit to the present moment. Still, the know-how transfer from laboratory mini-model to the real-size field ATIRV was a significant challenge. It required a precise control of kinematics and dynamics of the robotic vehicle. The propulsion and the braking needed to be properly managed and precisely monitored. A display and graphic presentation [14] of the actual values were important in fast monitoring process. Some descriptors in the process of designing the automated vehicle and from the SWOT analysis to study its faults and vulnerabilities are offered by Table4.

Table 4. SWOT analysis descriptors used in faults-and-vulnerabilities mapping of the application.

Strengths Weaknesses Opportunities Threats Controllable memory Storage volatility Deﬁning standards Buﬀer failure Remotely controlled Complex protocols Programming Program hacks Internet communication Potential privacy risks Extensive reach Data leaks

Development and research of automatic ATIRVs, using VR simulation and laboratory models, has significant importance in automation and robotics, Computer Aided Design/Computer Aided Manufacturing/Computer Aided Engineering (CAD/CAM/CAE) integrated systems, sensing and control, information systems technology, advanced software technology, intelligent systems and technologies, parallel and distributed computing technology, automotive engineering and transportation due to improvement of road and traffic safety, data processing techniques, and operational reliability in the coming years. In the present case, the proposed and tested ATIRV with an automatic guided driving program had a metal object detection system placed in the front part, a powertrain, and a steering system inside the vehicle’s structure to drive it optimally in a given scenario. The researched robot in the present study was a customized all-terrain intelligent guided vehicle that had the capacity to detect magnetic materials, but it could also be reconfigured for other specialized tasks. The ATIRV robot acted and drove independently and automatically on its own data as soon as the program was downloaded inside the electronic control unit. The electronic control unit was used for both powertrain and detecting system, which were interconnected. The electronic control unit drove the robotic vehicle and sustained the working process of the automated ATIRV. Compared to Havlik’s robotic tools for de-mining operations, the present work is still a progressing project, but there are some expectations for development.

Author Contributions: Conceptualization, F.C. and D.-L.B.; methodology, D.-L.B.; software, F.C.; validation, P.B., D.-L.B., and F.C.; formal analysis, D.-L.B.; investigation, P.B.; resources, P.B.; data curation, D.-L.B.; writing—original draft preparation, F.C.; writing—review and editing, D.-L.B.; visualization, D.-L.B.; supervision, P.B.; project administration, D.-L.B.; funding acquisition, P.B. All authors have read and agreed to the published version of the manuscript. Authorship must be limited to those who have contributed substantially to the work reported. Funding: This research project was co-funded by the European Social Fund. Acknowledgments: This paper was supported by the “Entrepreneurial competences and excellence research in doctoral and postdoctoral programs—ANTREDOC” project, co-funded by the European Social Fund. Conﬂicts of Interest: The authors declare no conﬂict of interest. Proceedings 2020, 63, 38 10 of 11

References

1. Aboulissane, B.; Haiek, D.; Bakkali, L.; Bahaoui, J. On the Workspace Optimization of Parallel Robots Based on CAD Approach. Procedia Manuf. 2019, 32, 1085–1092. [CrossRef] 2. Alejo, D.; Mier, G.; Marques, C.; Caballero, F.; Merino, L.; Alvito, P. SIAR: A Ground Robot Solution for Semi-autonomous Inspection of Visitable Sewers. In Advances in Robotics Research: From Lab to Market; Grau, A., Morel, Y., Puig-Pey, A., Cecchi, F., Eds.; Springer: Berlin/Heidelberg, Germany, 2020; Volume 132, pp. 275–296. 3. Andrei, L.; Băldean, D.; Borzan, A.I. Applied Measurements and Instrumentation for Improving Diagnostic Devices and Systems in Metropolitan Polluted Environments with Nitric and Carbon Oxides. In Proceedings of the 6th International Conference on Advancements of Medicine and Health Care through Technology, Cluj-Napoca, Romania, 17–20 October 2018; Vlad, S., Roman, N., Eds.; Springer: Berlin/Heidelberg, Germany, 2019; Volume 46, pp. 45–49. [CrossRef] 4. Băldean, D. Comparative Analysis of Three Distinct Porsche Models Concerning N.V.H. Aspects. In Proceedings of the 30th SIAR International Congress of Automotive and Transport Engineering, Craiova, Romania, 22–25 October 2019; Volume 1, pp. 209–213. [CrossRef] 5. Baldean, D.L.; Covaciu, F. Developing the communication of autonomous vehicles controlled with the aid of artificial intelligence for person and capital safety. In Safety of the Person and the Construction of Social Capital; Nechita Iancu, E.-A., Ed.; Universul Juridic: Arad, Romania, 2020; Volume 1, pp. 478–484. ISBN 978-606-39-0579-7. 6. Baldean, D.-L.; Covaciu, F.-A. Robotic Art in Creation and Development of Innovative Shapes and Programs for Automated Driven Cars with Artificial Intelligence. J. Soc. Media Inq. 2020, 2, 22–39. [CrossRef] 7. Bec, P.; Borzan, A.I.; Frunză, M.; Băldean, D.L.; Berindei, I. Study of Vulnerabilities in Designing and Using Automated Vehicles based on SWOT method for Chevrolet Camaro. In IOP Conference Series Materials Science and Engineering; IOP Publishing: Bristol, UK, 2020; Volume 898. 8. Borzan, A.I.; Băldean, D. Application of advanced engineering methods in studying a road traffic event between a 12-wheeler truck and a small tourism in a local junction from Cluj-Napoca. Univ. Pitesti. Sci. Bull. Automot. Ser. 2017, 27, 97–106. [CrossRef] 9. Borzan, A.-I.; Băldean, D. The Development of a New Interface for Intelligent Control of Energy Supply in Dynamic Environment with Process Digitization. Procedia Manuf. 2020, 46, 914–921. [CrossRef]

10. Cărăus, an, H.; Baldean, D.-L.; Borzan, A.-I. Analysis of Experimental Cross-Sectional Tests by Telemetry of Load Parameters in Relation to the Charge on a Mitsubishi Lancer. In National Bibliometric Instrument; IDSI: Chisinau, Moldova, 2018; Volume 51, pp. 217–220. Available online: https://ibn.idsi.md/sites/default/ ﬁles/imag_ﬁle/217-220_4.pdf (accessed on 10 June 2020).

11. Chereches, , I.A.; Baldean, D.-L.; Borzan, A.-I. Research of Intelligent Control of Injection Systems for Subaru Competition Car. In Proceedings of the 30th SIAR International Congress of Automotive and Transport Engineering, Proceedings of the SMAT 2019, Craiova, Romania, 23 October 2019; Dumitru, I., Covaciu, D., Racila, L., Rosca, A., Eds.; Springer: Berlin/Heidelberg, Germany, 2020; pp. 82–87. [CrossRef] 12. Haiek, D.; Aboulissane, B.; Bakkali, L.; Bahaoui, J. Optimal Trajectory Planning for Spherical Robot Using Evolutionary Algorithms. Procedia Manuf. 2019, 32, 960–968. [CrossRef] 13. Havlik, S. Robotic tools for de-mining and risky operations. In Using Robots in Hazardous Environments; Woodhead Publishing: Sawston, UK, 2011; pp. 327–352. 14. Jovrea, S.; Băldean, L.D.; Borzan, I.-A. Researching on-board display of essential information concerning technical conditions in operation and fuel-economy of a motor-vehicle in operation. In Science and Engineering; AGIR: Bucharest, Romania, 2017; Volume 31, Available online: https://stiintasiinginerie.ro/31-67 (accessed on 4 May 2017). 15. Oltean, S.-E. Mobile Robot Platform with Arduino Uno and Raspberry Pi for Autonomous Navigation. Procedia Manuf. 2019, 32, 572–577. [CrossRef] 16. Park, C.; Chung, S.; Lee, H. Vehicle-in-the-Loop in Global Coordinates for Advanced Driver Assistance System. Appl. Sci. 2020, 10, 2645. [CrossRef] 17. Yatimi, H.; Ouberri, Y.; Aroudam, E. Enhancement of Power Production of an Autonomous PV System Based on Robust MPPT Technique. Procedia Manuf. 2019, 32, 397–404. [CrossRef] 18. Thrun, S. Toward robotic cars. Commun. ACM 2010, 53, 99–106. [CrossRef] Proceedings 2020, 63, 38 11 of 11

19. Autonomous Robot. Available online: https://en.wikipedia.org/wiki/Autonomous_robot (accessed on 8 June 2020).

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional aﬃliations.

© 2020 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/).