A comprehensive performance evaluation of different mobile manipulators used as displaceable 3D printers of building elements for the construction industry

sustainability Article A Comprehensive Performance Evaluation of Different Mobile Manipulators Used as Displaceable 3D Printers of Building Elements for the Construction Industry Robert Guamán Rivera 1, Rodrigo García Alvarado 2and Alejandro Martínez-Rocamora 3 and Fernando Auat Cheein 1,* 1Department of Electronics Engineering, Universidad Técnica Federico Santa María, Valparaíso 1680, Chile; [email protected] 2Department of Design and theory of Architecture, Universidad de Bio-Bio, Concepción 1202, Chile; rgar[email protected] 3ArDiTec, Department of Architectural Constructions II, IUACC, Higher Technical School of Building Engineering, Universidad de Sevilla, Av. Reina Mercedes 4-a, 41012 Sevilla, Spain; r[email protected] *Correspondence: [email protected] Received: 23 March 2020; Accepted: 2 May 2020; Published: 27 May 2020   Abstract: The construction industry is currently technologically challenged to incorporate new developments for enhancing the process, such as the use of 3D printing for complex building structures, which is the aim of this brief. To do so, we show a systematic study regarding the usability and performance of mobile manipulators as displaceable 3D printing machinery in construction sites, with emphasis on the three main different existing mobile platforms: the car-like, the unicycle and the omnidirectional (mecanum wheeled), with an UR5 manipulator on them. To evaluate its performance, we propose the printing of the following building elements: helical, square, circular and mesh, with different sizes. As metrics, we consider the total control effort observed in the robots and the total tracking error associated with the energy consumed in the activity to get a more sustainable process. In addition, to further test our work, we constrained the robot workspace thus resembling real life construction sites. In general, the statistical results show that the omnidirectional platform presents the best results –lowest tracking error and lowest control effort– for circular, helicoidal and mesh building elements; and car-like platform shows the best results for square-like building element. Then, an innovative performance analysis is achieved for the printing of building elements, with a contribution to the reduction of energy consumption. Keywords: mobile manipulator; 3D printing; tracking trajectory 1. Introduction Construction automation (CA) is defined as the integration of intelligent machines (e.g., building robots and/or embedded and dedicated systems) [ 1 ], printing methods (such as 3D printing, concrete printing, contour crafting, D-Shape [ 2 ]), traditional construction methods [ 3 ] and advanced construction technologies (such as i) extrusion-based AM and ii)binder jetting) [4]. CA promises several benefits to the construction industry to enhance the productivity, quality, and sustainability of architectural practices and building construction [5]. Additive Manufacturing (AM) technologies allow the horizon of the 3D printing to be expanded in the construction sector. In this context, binder jetting AM is a technology that performs the selective deposition of a binder solution through a print nozzle onto a previously deposited powder layer [ 6 ]. Sustainability 2020,12, 4378; doi:10.3390/su12114378 www.mdpi.com/journal/sustainability Sustainability 2020,12, 4378 2 of 17 Furthermore, the additive manufacturing of concrete extrusion-based has recently employed in the construction industry. This technology involves extruding the cement-based material with nozzles of different sizes to build a layered structure model [7]. The technology of 3D printing in construction applications can be considered as a process where various materials are successively solidified layer upon layer by extrusion, to form solid models [ 8 ]. The major applications of 3D printing in large scale infrastructure are divided into three scenarios: D-Shape [ 9 ], contour crafting (CC) [ 10 ] and concrete printing [ 11 ]. The former, D-Shape, is based on binder injection and, despite its high effectiveness, it is only used for customized constructions [ 12 ]. For the second case, CC, it is a computer-controlled method with portability and cost improvements: it enables to build smooth surfaces in a short time [ 13 , 14 ]. Finally, concrete printing is very similar to CC, but it is associated with the development of components without formwork [15]. The success of 3D printing in CA depends on the quality of the materials. The printable mixture is similar and varies in composition from the traditional cement paste. For this reason, the printing materials must have appropriate rheological and compositional properties that allow: easy extrusion, strong adhesion between the printing layers, avoid the collapse of the structure and maintain the printing pattern during and after the material deposition process [ 4 ]. In this context, [ 16 ] presents an ultra-high performance concrete (UHPC) developed to accelerate the printing process and to improve the mechanical strength for layer-by-layer construction. One of the main motivations to consider UHPC is the high mechanical performance to build concrete-based structures, which is an alternative to more traditional construction methods (see [ 17 – 19 ] for further details). Furthermore, to achieve structural integrity, durability, reliability, and robustness without any support structures. Thus, ref. [ 20 ] presents a 3D printing based on Engineered Cementitious Composites (ECC), whose advantages are sustainable mix design, rheology control, and long-term durability of the 3D printing. Additive Manufacturing (AM) technology using the extrusion-based method has been improving with the integration of robotic systems; specifically, a robot manipulator handling the extrusor and controlling the material deposition [ 21 – 23 ]. For example, in [ 24 , 25 ] it is shown the customisation of a robotic arm in the construction industry, aimed at printing concrete walls of different geometries. The main advantage of using robot manipulators is their workload capacity, their accuracy in repetitive tasks and the flexible programming [ 26 , 27 ]. Nevertheless, robot manipulators in construction applications are mainly used as fixed machinery, without interacting with the environment or moving within the construction site [16]. The construction industry uses robots for assembly and disassembly of various components. In particular, ref. [ 28 ] identifies four stages for assembling building components: (i) no assembly components, (ii) assembly of a large element formed with several small components, (iii) placement of components in the final position, and (iv) assembly of non-printed external components. In this context, ref. [ 29 ] presents a Robotic Prefabrication System (RPS) that allows the automatic disassembly of a prefabricated structure and determinates the needs and gaps in knowledge in the current prefabrication. Moreover, as a mechanical tools, robotic arms have advantages that attract the attention of the construction industry, such as their dexterity and their reachability. However, such advantages also restrict the scale size of the construction: when fixed in the ground, a robot manipulator with an extrusor (used for printing) can only print elements that fall within its workspace. Otherwise, the robot has to be manually displaced [30–33]. To overcome the later problem, in this work we propose to study the usability of a robotic arm mounted on a mobile platform (also robotic), to cover bigger areas of the construction site and to print convex geometries (for example, the walls of a room). By integrating a mobile platform to the robot manipulator used as 3D printer, we increase its dexterity, its workspace and its capabilities. However, since we use the same robot manipulator, we change the mobile platform to find which one is the most appropriate to be used in the construction site. In particular, we test the following configurations: unicycle, car-like and four wheeled omnidirectional with mecanum wheels [ 34 , 35 ]. We test the performance of the mobile manipulator mounted on the three different platforms in terms Sustainability 2020,12, 4378 3 of 17 of error and effort, when printing several types of building elements. For the remainder of this work, we will refer to the robotic platform as the mobile manipulator. The selection of proper machine and printing strategy according to the building pieces printed is crucial for a successful efficiency and versatility of this emerging technology for 3D-printed construction with robotic systems. In this context, the efficient use of robotic platforms is related to their operational power consumption [ 36 ]. When the robot performs a printing task, the power consumption varies significantly during the operation of the robot, and comprise the actuation systems that will interact with the environment and the construction process. For example, the use of energy resources of the mobile manipulator depends on the interaction of the platform with the environment and the motion it executes. For this reason, their energy use is governed by the irregularity of the surface of the construction environment, the complexity of the print model, and the payload it supports. In this work, we relate the energy consumption to the cost function in terms of minimum trajectory tracking error with the motion speeds of the mobile manipulator. As previously stated, this work is focused on studying the usability of mobile manipulator platforms as 3D printing machinery in construction sites, with emphasis in the different existing mobile platforms, in order to provide a novel review and suggest adequate methodologies for buildings construction. Previous work has been done on materials, machines and/or specific printing experiments, but lack of rigorous studies and general approaches about printing procedures with mobile robot platforms. Movement and operation of robots are complex tasks to be combined in a printing process of large pieces, and building execution requires a diversity of elements to be printed in different locations, then to test and define proper strategies is essential to develop the construction with robots. In this context, [ 8 , 37 ] exposes strategies of mobile systems for printing building elements but do not study the toolpath in relation to the machine type and piece design. Such procedure is relevant to develop the printing methodology and define effective approaches and equipments. This work is organized as follows: Section 2shows the trajectory profiles in the construction environment, the mathematical derivation of the mobile manipulator model, the strategy followed by the mobile manipulator trajectory tracking controller, and the metrics followed to find mobile platform performance. Section 3presents the results using the mobile manipulator with the three different mobile platforms previously introduced for different building scenarios. Section 4shows a discussion about the tasks performed by the mobile manipulator in the construction environments. Finally, the conclusions are shown in Section 5. 2. Materials and Methods Figure 1shows the general scheme of the architecture proposed in this work for evaluating the performance of mobile manipulators in the 3D printing case of building elements. To the left, we have two reference profiles: circular and square, in two different views: 2D (two dimensional) and 3D. Such profiles represent the building elements to be printed. When sent to the manipulator, the building elements are converted into path references to be tracked by the end effector (where the extrusor is located) of the manipulator. We consider three mobile platforms where the manipulator is going to be mounted: unicycle type, car-like type (Ackerman) and omnidirectional type (four wheeled with mecanum wheels). The trajectory references are then transformed into motion control commands through a linear algebra controller to ensure that the system mobile manipulator behaves as expected: its end effector—to which the nozzle is attached—moves describing the building element to be printed. In the following subsections, each part of Figure 1is explained in detail. Sustainability 2020,12, 4378 4 of 17 Omnidirectional UnicycleCarlike Linear Algebra Controller 2D profile 3D profile 2D profile 3D profile 0 0 0.1 z [m] 2 0.2 y [m] 0 0.3 4 x [m] 2 4 66 Desired Actual 3 0.8 2 y [m] 3 x [m] 12 0.9 z [m] 1 1 Desired Actual 4 8 y [m] 6 x [m] 6 4 0.5 28 z [m] 1 Desired Actual 0.2 0 0.4 4 1 0.6 z [m] x [m] y [m] 2 2 0.8 30 Desired Actual Reference profile ResultMobile manipulator Figure 1. Construction process of a trajectory profile developed by a mobile robotic system. Mobile manipulator with three configurations for robot mobile. The robotic structure is formed of base mobile (unicycle, car-like and omnidirectional) and manipulator. 2.1. Robot Manipulator The robotic arm used to perform 3D printing of the building elements is the UR5, manufactured by Universal Robots. This manipulator has six degrees of freedom and a control unit that provides basic joints control as well as a companion computer-computational system in charge of data processing and broadcasting. It supports a payload of 5 kg; its maximum speed is π rad/s and its repeatability is ±0.1 mm. Table 1shows the main features of this manipulator. Table 1. Robot manipulator: UR5 main features. Feature Value Weight 18.4 kg Payload 5 kg Reach 850 mm Joint Ranges ±2π Speed πrad/s I/O power supply 24 V 2 A Communication TCP/IP 100 Mbit:IEEE 802.3u, 100BASE-TX Programing Polyscope graphical user interface Temperature The robot can work in a temperature range of 0–50 degrees One of the main advantages of this robot is that it offers low-level programing with high cycle time. The robot can be modelled using Denavit-Hartenberg parameters or geometrical model and it has been used in novel applications in different industries [ 38 ]. Although we used the UR5 robotic arm in this work, the procedure followed and presented herein can be extended to other manipulators. 2.2. Mobile Robot Table 2shows the kinematic models in continuous and discrete-time of the non-holonomic (i.e., unicycle and carlike) and holonomic (i.e., omnidirectional robot) platforms studied in this work. We chose such three mobile platforms since they are the most used ones as reported in the literature [39–41] . However, as also stated for the UR5 case, if needed, the procedure presented in this work can be adapted to other types of platforms. As it is shown in Table 2, the three robots have three degrees of freedom, named as x , y and ψ , which apply to rotation and translation of the robot in the plane, considering only planar terrains. In the case of the unicycle robot, it can turn around its control point [ 42 ]; however, the car-like model has to follow a circle based path [ 43 ], making it impossible to turn without displacing [ 44 ]. On the other hand, the omnidirectional vehicle can displace at any direction in the plane. More information regarding the kinematic (and dynamic) constraints of each robotic type can be found in [45–47]. Sustainability 2020,12, 4378 5 of 17 Table 2. Kinematic model of each mobile platform. Model Continuos Time Discrete Time Unicycle [39] ˙ x=µcos ψ−aωsin ψx(n+1)=x(n)+µ(n)cos ψ(n)−aω(n)sin ψ(n) ˙ y=µsin ψ+aωcos ψy(n+1)=y(n)+µ(n)sin ψ(n)+aω(n)cos ψ(n) ˙ ψ=ω ψ(n+1)=ψ(n)+ω(n) Car-like [40] ˙ x=µcos ψ−aωsin ψx(n+1)=x(n)+µ(n)cos ψ(n)−aω(n)sin ψ(n) ˙ y=µsin ψ+aωcos ψy(n+1)=y(n)+µ(n)sin ψ(n)+aω(n)cos ψ(n) ˙ ψ=µ Ltan (δ)ψ(n+1)=ψ(n)+µ(n) Ltan δ(n) Omnidirectional [41] ˙ x=1 4γv1−1 4βv2+1 4γv3+1 4βv4x(n+1)=x(n)+1 4γv1(n)−1 4βv2(n)+1 4γv3(n)+1 4βv4(n) ˙ y=1 4βv1−1 4γv2+1 4βv3+1 4γv4y(n+1)=y(n)+1 4βv1(n)−1 4γv2(n)+1 4βv3(n)+1 4γv4(n) ˙ ψ=−v1 4(l+d)−v2 4(l+d)+v3 4(l+d)+v4 4(l+d)ψ(n+1)=ψ(n)−v1(n) 4(l+d)−v2(n) 4(l+d)+v3(n) 4(l+d)+v4(n) 4(l+d) Sustainability 2020,12, 4378 6 of 17 In Table 2, µ and ω are the linear and angular velocities of the unicycle and car-like robot; a is the distance between the centre of the platform and the centre of mass in global coordinates; δ is the heading of the car-like; L is the length between axles of the car-like robot and (l+d) is the distance from the centre of the wheel to the centre of mass of the omnidirectional robot; v1 , v2 , v3 , v4 are wheel velocities of the omnidirectional robot. Furthermore, γ is defined by (cos ψ−sin ψ) and β is defined by (sin ψ+cos ψ). Suffix nrepresents sampling time. 2.3. Mobile Manipulator As stated in Section 1, when the robot manipulator is attached to one of the mobile platforms mentioned in Section 2.2, the combined system becomes a mobile manipulator. Its kinematic is the result of also combining the manipulator kinematics with the mobile robot kinematics. Hence, we use the Denavit-Hartenverg convention to derive the mobile manipulator kinematic model, following the guidelines previously published by [38]. Table 3shows the kinematic model of the robotic arm mounted on the three different mobile platforms. For unicycle and car-like configurations, linear and angular velocity defines the motion of the mobile platform. The orientation of the unicycle robot is defined by angular rotation. However, in the car-like configuration, the rotation is a function of the linear velocity and the length of the mobile platform. For the omnidirectional robot, rotation and linear velocity are a function of the angular velocity described by the wheels. Table 3. Kinematic model of the mobile manipulator. Mobile Manipulator Kinematic Model Unicycle [˙ xee ˙ yee ˙ zee]T=J[µ ω ˙ θ1˙ θ2˙ θ3˙ θ4˙ θ5]T Car-like [˙ xee ˙ yee ˙ zee]T=J[µµ L˙ θ1˙ θ2˙ θ3˙ θ4˙ θ5]T Omniderectional [˙ xee ˙ yee ˙ zee]T=J[v1v2v3v4˙ θ1˙ θ2˙ θ3˙ θ4˙ θ5]T where, J=∂hee ∂f(x,y,ψ,θ1, ..., θ5)(1) and hee = [xee yee zee]T is the position of the end-effector –with respect to some global reference frame–, and [θ1θ2θ3θ4θ5]T are its joint angles. Furthermore, [˙ xee ˙ yee ˙ zee]T is the velocity of the end-effector; ˙ θ1 , ˙ θ2 , ˙ θ3 , ˙ θ4 and ˙ θ5 are angular velocities of the joints. Figure 2shows the manipulator mounted on three mobile platforms (unicycle, car-like and omnidirectional with four mecanum wheels). Angles and velocities are represented according to a global reference frame G , R is the local coordinate of the mobile robot, M represents the local coordinate of the manipulator UR5 and the coordinate of the end-effector is Oe . The manipulator links are represented by d1 , a2 , a3 , d4 , and d5, and the height of the mobile platform is h1. Sustainability 2020,12, 4378 7 of 17 x1 y1 z1 x y z x2 y2z2x4 y4 z4 M x3 y3z3 x5 y5 z5 xee yee zee xG yG zG G d 1 a2 a3 d 5 d 4 d 6 Oe R h 1 (a) x2 y2z2 x4 y4 z4 x3 y3z3 x5 y5 z5 xee yee zee a2 a3 d 5 d 4 d 6 Oe xG yG zG G x1 y1 z1 x y z M R d 1 h 1 (b) x y z R xG yG zG G x2 y2z2 x4 y4 z4 x3 y3z3 x5 y5 z5 xee yee zee a2 a3 d 5 d 4 d 6 Oe y1 z1 Mx1 d 1 h 1 (c) Figure 2. Mobile manipulator model: robotic arm with six degree freedom mounted in mobile platform; ( a ) shows the unicycle case, ( b ) shows the car-like case and ( c ) the omnidirectional case with four mecanum wheels. 2.3.1. Workspace Restriction Since this work is focused on analysing the usability of mobile manipulators in construction environments, it becomes necessary to also analyse the workspace surrounding the platform due to the fact that its motion is constrained by the environment layout, the terrain and the task restrictions, the later with the aim of guaranteeing the effectiveness and safety of the operation. In our work, the first constraint impossed to the system is the restriction of the workspace, by two times the maximum extension of the robot manipulator. Thus, it limits the displacement of the platform by allowing it no more than one time the robot manipulator extension, to ensure the safety of the operation. Other criteria might apply since it is to be noted that the platform does not move freely in a construction site. In the end, the idea is to consider the design requirements and shapes of the building elements to evaluate the most appropriate workspace, to thus avoid excuting tasks at the maximum point of operation, or to invade an area beyond the printed building element [ 19 ]. Figure 3 summarizes the restrictions considered in this work for all the building elements studied, where the solid red line represents the outer limit imposed on the mobile platform, this limit may not exceed twice the maximum length of the robotic arm; the solid blue line is the inner limit and cannot exceed the maximum arm length. Sustainability 2020,12, 4378 8 of 17 Figure 3. Constraints of the workspace. The solid red line represents the maximum limit of operation of the mobile platform and the minimum limit is in a solid blue line. Therefore, the mobile robot should not trespass both squares. 2.4. Building Elements The construction industry uses complex building element profiles to build houses and buildings. These surfaces result in high profile projects or iconic architecture designs. The main challenges of building element profiles are their complex geometries, size and repeatability of the task, which can affect the performance of the robotic systems [ 48 , 49 ]. Therefore the need of using mobile manipulators. We generate four trajectory profiles following the guidelines presented in [ 26 ]. Such profiles are then converted into reference trajectories fed to the mobile manipulator, parametrized in time and discretized with sampling time of 0.1 s (however, other sampling criterion might apply). The building elements studied in this work are: circle, helical, mesh and square; which are also depicted in Figure 4. M e s h Circle Square Hel i c al Figure 4. Building elements proposed in this work. We tested three different sizes of the above building element: small (length = 1 m, width = 1 m, height = 1.2 m); medium (length = 2 m, width = 2 m, height = 1.2 m); and large (length = 5 m, width = 5 m , height = 1.2 m), following the guidelines presented in [ 50 , 51 ]. Such building elements are uniform, except for the corners in square building elements, where the change is abrupt and might affect the orientation of the mobile platform. For the circle and helical case, the diameter proposed was of 1 m (small size), 2 m (medium size) and 4 m (large size). The generation of the building elements is divided into 2D layers to generate 3D model geometries. Each building element is represented by circular and square geometries. Additionally, for printing purposes, we added the following constraints: printing speed between 0.01 and 1 ms −1 [ 52 ]; maximum Sustainability 2020,12, 4378 9 of 17 mobile platform speed of 0.7 ms −1 ; and we limited the maximum angles of the joints to − 2 π to 2 π rad [53]. 2.5. Motion Control In this work, we propose a trajectory tracking algorithm for the mobile manipulator to ensure 3D printing of building elements at constant speed. To this end, a trajectory tracking algorithm is implemented using the controller proposed by Scaglia et al. [ 54 ]. This controller guarantees velocity regulation between the end effector and the printing task. The kinematic model presented in Table 3can be expressed as follows: ˙ h(t)=fΓ,˙ θ1,˙ θ2,˙ θ3,˙ θ4,˙ θ5(2) The control variables of the mobile platform are represented by Γ . We then defined Γ for each mobile platform and replaced it in Equation (2), where Γ=µ , ω represents the control variables of the unicycle robot, Γ=µ , µ L is defined for a car-like robot and Γ=v1 , v2 , v3 , v4 represents the control signals of the omnidirectional robot. The continuous system shown in Equation (2) can be rewritten discretized using an Euler approach as shown below, for the three models: h(n+1)=h(n)+Z(n+1)To nTo fΓ,˙ θ1,˙ θ2,˙ θ3,˙ θ4,˙ θ5dt (3) h(n+1)∼ =h(n)+T0fΓ(n),θ1(n),θ2(n),θ3(n),θ4(n),θ5(n)(4) fΓ(n),θ1(n),θ2(n),θ3(n),θ4(n),θ5(n)=J(n)           Γ(n) θ1(n) θ2(n) θ3(n) θ4(n) θ5(n)           (5) where J(n) is the discrete Jacobian matrix of J from Table 3at each sampling time. The values of h(t) at discrete time t=nTo , where To is the sampling period, and n∈{0, 1, 2, . . .} , are denoted as h(n)=hxee(n),yee(n),zee(n)iT and h(n+1)=hxee(n+1),yee(n+1),zee(n+1)iT . The kinematic model of the mobile manipulator, is defined by:     xee(n+1)−xee(n) To yee(n+1)−yee(n) To zee(n+1)−zee(n) To     =J(n)U(n)(6) where U(n)=hΓ(n)θ1(n)θ2(n)θ3(n)θ4(n)θ5(n)iT. The proposed control law is defined by: Uc(n)=J+ (n)     xeed(n+1)−kxxeed(n)−xee(n)−xee(n) yeed(n+1)−kyyeed(n)−yee(n)−yee(n) zeed(n+1)−kzzeed(n)−zee(n)−zee(n)      , (7) where J+ (n) is the Jacobian pseudo-inverse matrix of the mobile manipulator, the position of the end-effector is defined by [xee(n)yee(n)zee(n)]T framed in a global reference system, and K= [kxkykz)]T is the set of tuning parameters. The desired path of the end-effector is given Sustainability 2020,12, 4378 16 of 17 19. Furet, B.; Poullain, P.; Garnier, S. 3D printing for construction based on a complex wall of polymer-foam and concrete. Addit. Manuf. 2019,28, 58–64. [CrossRef] 20. Li, V.C.; Bos, F.P.; Yu, K.; McGee, W.; Ng, T.Y.; Figueiredo, S.C.; Nefs, K.; Mechtcherine, V.; Nerella, V.N.; Pan, J.; et al. On the emergence of 3D printable Engineered, Strain Hardening Cementitious Composites (ECC/SHCC). Cem. Concr. Res. 2020,132, 106038. [CrossRef] 21. Dharmawan, A.G.; Sedore, B.W.C.; Foong, S.; Soh, G.S. An agile robotic system mounted on scaffold structures for on-site construction work. Constr. Robot. 2017,1, 15–27. [CrossRef] 22. Bard, J.; Cupkova, D.; Washburn, N.; Zeglin, G. Robotic concrete surface finishing: A moldless approach to creating thermally tuned surface geometry for architectural building components using Profile-3D-Printing. Constr. Robot. 2018,2, 53–65. [CrossRef] 23. Dörfler, K.; Hack, N.; Sandy, T.; Giftthaler, M.; Lussi, M.; Walzer, A.N.; Buchli, J.; Gramazio, F.; Kohler, M. Mobile robotic fabrication beyond factory conditions: Case study Mesh Mould wall of the DFAB HOUSE. Constr. Robot. 2019,3, 53–67. [CrossRef] 24. Reinhardt, D.; Titchkosky, N.; Bickerton, C.; Watt, R.; Wozniak-O’Connor, D.; Candido, C.; Cabrera, D.; Page, M.; Bohnenberger, S. Towards onsite, modular robotic carbon-fibre winding for an integrated ceiling structure. Constr. Robot. 2019,3, 23–40. [CrossRef] 25. Bos, F.; Wolfs, R.; Ahmed, Z.; Salet, T. Additive manufacturing of concrete in construction: Potentials and challenges of 3D concrete printing. Virtual Phys. Prototyp. 2016,11, 209–225. [CrossRef] 26. Davtalab, O.; Kazemian, A.; Khoshnevis, B. Perspectives on a BIM-integrated software platform for robotic construction through Contour Crafting. Autom. Constr. 2018,89, 13–23. [CrossRef] 27. Mechtcherine, V.; Nerella, V.N.; Will, F.; Näther, M.; Otto, J.; Krause, M. Large-scale digital concrete construction–CONPrint3D concept for on-site, monolithic 3D-printing. Autom. Constr. 2019 ,107, 102933. [CrossRef] 28. Perrot, A.; Amziane, S. 3D Printing in Concrete: General Considerations and Technologies. 3D Print. Concr State Art Chall. Digit. Constr. Revolut. 2019, 1–40. [CrossRef] 29. Kasperzyk, C.; Kim, M.K.; Brilakis, I. Automated re-prefabrication system for buildings using robotics. Autom. Constr. 2017,83, 184–195. [CrossRef] 30. Perrot, A.; Rangeard, D.; Pierre, A. Structural built-up of cement-based materials used for 3D-printing extrusion techniques. Mater. Struct. 2016,49, 1213–1220. [CrossRef] 31. de Soto, B.G.; Agustí-Juan, I.; Hunhevicz, J.; Joss, S.; Graser, K.; Habert, G.; Adey, B.T. Productivity of digital fabrication in construction: Cost and time analysis of a robotically built wall. Autom. Constr. 2018 , 92, 297–311. [CrossRef] 32. Ngo, T.D.; Kashani, A.; Imbalzano, G.; Nguyen, K.T.; Hui, D. Additive manufacturing (3D printing): A review of materials, methods, applications and challenges. Compos. Part B Eng. 2018,143, 172–196. [CrossRef] 33. Buchanan, C.; Gardner, L. Metal 3D printing in construction: A review of methods, research, applications, opportunities and challenges. Eng. Struct. 2019,180, 332–348. [CrossRef] 34. Wang, C.; Liu, X.; Yang, X.; Hu, F.; Jiang, A.; Yang, C. Trajectory tracking of an omni-directional wheeled mobile robot using a model predictive control strategy. Appl. Sci. 2018,8, 231. [CrossRef] 35. Pappalardo, C.M.; Guida, D. Forward and Inverse Dynamics of a Unicycle-Like Mobile Robot. Machines 2019,7, 5. [CrossRef] 36. Martínez-Rocamora, A.; García-Alvarado, R.; Casanova-Medina, E.; González-Böhme, L.F.; Auat-Cheein, F. Parametric Programming of 3D Printed Curved Walls for Cost-Efficient Building Design. J. Constr. Eng. Manag. 2020,146, 04020039. [CrossRef] 37. Giftthaler, M.; Sandy, T.; Dörfler, K.; Brooks, I.; Buckingham, M.; Rey, G.; Kohler, M.; Gramazio, F.; Buchli, J. Mobile robotic fabrication at 1: 1 scale: The in situ fabricator. Constr. Robot. 2017,1, 3–14. [CrossRef] 38. Kebria, P.M.; Al-Wais, S.; Abdi, H.; Nahavandi, S. Kinematic and dynamic modelling of UR5 manipulator. In Proceedings of the 2016 IEEE International Conference on Systems, Man, and Cybernetics (SMC), Budapest, Hungary, 9–12 October 2016; pp. 4229–4234. 39. Vallejo-Alarcón, M.; Castro-Linares, R.; Velasco-Villa, M. Unicycle-type robot & quadrotor leader-follower formation backstepping control. IFAC-PapersOnLine 2015,48, 51–56. 40. Moreno, J.; Clotet, E.; Lupiañez, R.; Tresanchez, M.; Martínez, D.; Pallejà, T.; Casanovas, J.; Palacín, J. Design, implementation and validation of the three-wheel holonomic motion system of the assistant personal robot (APR). Sensors 2016,16, 1658. [CrossRef] [PubMed] Sustainability 2020,12, 4378 17 of 17 41. Li, X.; Zell, A. Motion control of an omnidirectional mobile robot. In Informatics in Control, Automation and Robotics; Springer: Berlin/Heidelberg, Germany, 2009; pp. 181–193. 42. Kamel, M.A.; Zhang, Y. Decentralized leader-follower formation control with obstacle avoidance of multiple unicycle mobile robots. In Proceedings of the 2015 IEEE 28th Canadian Conference on Electrical and Computer Engineering (CCECE), Halifax, NS, Canada, 3–6 May 2015; pp. 406–411. 43. Song, Z.; Ren, H.; Zhang, J.; Ge, S.S. Kinematic analysis and motion control of wheeled mobile robots in cylindrical workspaces. IEEE Trans. Autom. Sci. Eng. 2015,13, 1207–1214. [CrossRef] 44. Raj, J.; Raghuwaiya, K.; Vanualailai, J.; Sharma, B. Navigation of Car-Like Robots in Three-Dimensional Space. In Proceedings of the 2018 5th Asia-Pacific World Congress on Computer Science and Engineering (APWC on CSE), Nadi, Fiji, 10–12 December 2018; pp. 271–275. 45. Patle, B.; Pandey, A.; Parhi, D.; Jagadeesh, A. A review: On path planning strategies for navigation of mobile robot. Def. Technol. 2019,15, 582–606. [CrossRef] 46. Rathinam, S.; Manyam, S.G.; Zhang, Y. Near-Optimal Path Planning for a Car-Like Robot Visiting a Set of Waypoints With Field of View Constraints. IEEE Robot. Autom. Lett. 2019,4, 391–398. [CrossRef] 47. Sun, Z.; Dai, L.; Liu, K.; Xia, Y.; Johansson, K.H. Robust MPC for tracking constrained unicycle robots with additive disturbances. Automatica 2018,90, 172–184. [CrossRef] 48. Tahmasebinia, F.; Niemelä, M.; Ebrahimzadeh Sepasgozar, S.; Lai, T.; Su, W.; Reddy, K.; Shirowzhan, S.; Sepasgozar, S.; Marroquin, F. Three-Dimensional Printing Using Recycled High-Density Polyethylene: Technological Challenges and Future Directions for Construction. Buildings 2018,8, 165. [CrossRef] 49. Tomé, A.; Vizotto, I.; Valença, J.; Júlio, E. Innovative Method for Automatic Shape Generation and 3D Printing of Reduced-Scale Models of Ultra-Thin Concrete Shells. Infrastructures 2018,3, 5. [CrossRef] 50. Bogue, R. 3D printing: The dawn of a new era in manufacturing? Assem. Autom. 2013 ,33, 307–311. [CrossRef] 51. Kietzmann, J.; Pitt, L.; Berthon, P. Disruptions, decisions, and destinations: Enter the age of 3-D printing and additive manufacturing. Bus. Horiz. 2015,58, 209–215. [CrossRef] 52. Izard, J.B.; Dubor, A.; Hervé, P.E.; Cabay, E.; Culla, D.; Rodriguez, M.; Barrado, M. Large-scale 3D printing with cable-driven parallel robots. Constr. Robot. 2017,1, 69–76. [CrossRef] 53. Universal Robots. 2020. Available online: https://www.universal-robots.com/es/ (accessed on 20 April 2020). 54. Scaglia, G.; Montoya, L.Q.; Mut, V.; di Sciascio, F. Numerical methods based controller design for mobile robots. Robotica 2009,27, 269–279. [CrossRef] 55. Scaglia, G.; Serrano, E.; Rosales, A.; Albertos, P. Linear interpolation based controller design for trajectory tracking under uncertainties: Application to mobile robots. Control Eng. Pract. 2015 ,45, 123–132. [CrossRef] 56. Deepyaman, M.; Ayan, A.; Mithun, C.; Amit, K.; Ramdoss, J. Tuning PID and PI λ D µ controllers using the integral time absolute error criteria. In Proceedings of the 4th International Conference on Information and Automation for Sustainability ICIAFS, Colombo, Sri Lanka, 12–14 December 2008; pp. 457–462. 57. Bos, F.P.; Ahmed, Z.Y.; Wolfs, R.J.; Salet, T.A. 3D printing concrete with reinforcement. In High Tech Concrete: Where Technology and Engineering Meet; Springer: Cham, Switzerland, 2018; pp. 2484–2493. c 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/).