Robotics Inspired Renewable Energy Developments: Prospective Opportunities and Challenges
The domain of Robotics is a good partner of renewable energy and is becoming critical to the sustainability and survival of the energy industry Moreso in East Africa & Horn of Africa as well. The multi-disciplinary nature of robots offers precision, repeatability, reliability, productivity and intelligence, thus rendering their services in diversified tasks ranging from manufacturing, assembling, and installation to inspection and maintenance of renewable resources. My paper explores applications of real robots in one feasible renewable energy domain; solar.
In each case, existing state-of-the-art innovative robotic systems are investigated that have the potential to create a difference in the corresponding renewable sector in terms of reduced set-up time, lesser cost, improved quality, enhanced productivity and exceptional competitiveness in the global market. Instrumental opportunities and challenges of robot deployment in the renewable sector are also discussed with a brief case study of Saudi Arabia. It is expected that the wider dissemination of the instrumental role of robotics in renewable energy will contribute to further developments and stimulate more collaborations and partnerships between professionals of robotics and energy communities.
INDEX TERMS Applied robotics, automation in renewable energy, mobile robots, robotic manipulators, solar PV module, wind turbines.
I. INTRODUCTION
Human civilization perpetually depends on the energy that has become a fundamental entity behind social, scientific and economic developments. Thanks to advancements in various technological avenues, reliance and dependence on fossil fuel are getting reduced due to limited availability, growing needs and environmental concerns. Therefore, since the last two decades, research and innovation in the renewable energy sector have attracted the utmost attention of the scientific and engineering communities. Primarily driven with economic growth and climate mitigation, it is anticipated that 100% of renewable energy may be available by 2050. This huge target can be achieved if and only if novel and innova- tive development, adaptation, commercialization and deploy- ment technologies are presented by the scientific community to strengthen further the applied research associated with the renewable energy sector. Consequently, there is an immense need to critically analyze the performance and via- bility of the existing processes to automate these using recent cutting edge technologies.
Robotics is an applied domain whose radius of applications is getting incredibly wider owing to its multi-disciplinary nature. Robotics engineering is now being acknowl- edged as a dedicated branch of engineering. Robots are extensively used in agriculture, food, medical, cognition, nuclear , space , aerospace, under-water, industrial, oil and gas, textile and in other tracking,applications. International Federation of Robotics (IFR) is considered as a primary resource of providing data on the worldwide use of robots. IFR classifies the robots into two broader categories; industrial robots and service robots. In autumn 2019, IFR released the latest statistics of robot and Section IV respectively classify robots for solar and wind sectors and present recent and prominent state-of-the- art developments. Section V deals with the contributions of robots in other domains of renewable energy, namely hydro energy and bio-energy. Section VI briefly outlines the chal- lenges and explores further opportunities for robotics in the renewable energy sector. A brief case study of Kingdom of Saudi Arabia (KSA) is presented in Section VII. Finally, Section VIII comments on the conclusion and highlights the potential benefits of this comprehensive review stating a 6% increase in global robot installations in 2018 with a record sales value of 16.5 billion USD. It is reported that a greater trend toward automation is the primary reason for the tremendous increase in the demand for industrial robots particularly since 2010. The speculated average yearly growth rate of industrial robots during 2019-2021 is 14%, with new installations of around 2.1 million setups round the globe. The estimated annual supply of industrial robots with a forecast in the coming years are presented in Fig. 1. Automo- tive, electrical/electronics and metal/machinery sectors are the top three industries w.r.t. deployment of industrial robots. Statistics for robot usage in the renewable energy sector is not available as of to date.
This paper is related to applications of robotics in the renewable energy domain. The relationship between robotics and renewable energy can be described in two folds; renew- able energy resources can be used to meet the power require- ments of the robots while on the other side, robots find enormous potential in renewable energy technologies. The former aspect is reviewed in detail by the author in. It is highlighted that the power system of a robot can be based on solar, wind and biological energy. In contrast, the present review deals with the later aspect. Robots have the potential to completely transform the traditional methods in the renewable energy sector. They can precisely perform common industrial tasks like machine tending, grasping, cutting, drilling, pol- ishing, painting, welding, assembling, palletizing, packing, moving, cleaning, sorting in addition to automating assembly lines and pick and place operations. Most of these operations are essentially required in the renewable energy industry, e.g. during manufacturing, assembling, installation, inspection and maintenance of panels and wind turbines. In combination with computer vision , image processing and control, robots offer intelligent and automated solutions to improve quality and enhance productivity with adaptability and efficiency at reduced costs and thus demonstrate a central role in making renewable energy resources more competitive.
KEY CONCEPTS BEHIND ROBOTIC MANIPULATORS Robots can be generally classified into manipulator-type robots and mobile robots. The former category is more common in the renewable energy sector. Understanding of the fundamental concepts behind robotic manipulators is important to adapt a general-purpose robot for a specific application. These concepts include but are not limited to; kinematics, dynamics, control, trajectory planning, cost, workers’ safety, ease in Operation and Maintenance (O&M), etc. Some of these concepts are detailed below:
FIGURE 2. Robot categories (a) Serial stage (b) Parallel Kinematic
Manipulator (PKM).
A. KINEMATICS AND DYNAMICS
A robotic manipulator can be kinematically based on serial or parallel mechanisms. The serial robots offer a significant percentage of robotics-based solutions in the renewable energy sector (See Fig. 2a). However, the robots with Parallel Kinematic Manipulator (PKM) structure (Fig. 2b) have been recently introduced in the energy sector.
An example of a platform centered on a 6-DOF (Degree Of Freedom) serial link articulated robotic manipulator is shown in Fig. 3. The framework named as AUTonomous Articulated Robotic Educational Platform (AUTAREP) has been devel- oped for educational and research purposes. The actuation system consists of six precise DC geared servo motors while the sensing system comprises of position encoders, a force- sensing resistor and an on-board camera. Applications of the platform implementing frequently encountered industrial tasks, e.g. pick and place and sorting are reported in while software and hardware architectures of the platform are presented in.
On the other hand, PKM is an ideal choice for light-weight applications with the robotic end-effector offering low inertia and high stiffness and high payload capacity. However, the workspace envelope of a PKM is relatively smaller when compared to serial manipulators. A popular example of PKM is FlexPicker, also called as ‘Delta’ robot. The robot accounts for 3-DOF configuration restricted to move only in translation.
One of the preliminary steps to realize a robotic system in an applied context is the modeling of its kinematics and dynamics behavior. Modeling can be based on Denavit- Hatengerg (D-H) parameters or other representations, including Hayati-Roberts (H-R), screw theory, geometrical, Lie Algebra, etc. A detailed review of the modelling of robotic manipulators is reported in. The forward kinematics of the AUTAREP manipulator is derived in, while the inverse kinematic model is reported in. Unlike serial manipulators, the direct solution in PKM cannot be derived analytically. For a PKM, a vector loop equation is formulated for each limb considering a closed-loop kinematic chain.
The extension of a robot’s kinematics is the development of dynamic models mainly used for acceleration analysis. The dynamics equations can be formulated based on several methods like Newton-Euler, Euler-Lagrange, recur- sive Lagrange, D’Alembert principle and Kane’s equations. The first two approaches are more commonly followed. The dynamic model of the AUTAREP manipulator based on the Euler-Lagrange method is reported in. For PKM, explicit equations representing system dynamics are complicated due to the closed kinematic chain in the manipulator. One way to build a computationally inexpensive model is to use the principle of virtual work.
B. CONTROL
Analysis and investigation of robotic manipulators in an applied context has highlighted the immense need for com- plex strategies for control and dexterity. Systematic reviews on current and emergent control strategies for robotic manipulators are reported in. Since the last three decades, industrial processes have been classically controlled based on linear control laws. Proportional Integral Derivative (PID) implementation of AUTAREP is reported in while Linear Quadratic Regulator is presented in [86]. The research community has lately applied advanced control
strategies based on modern and nonlinear control laws on multi-DOF robotic manipulators to deal with uncertain parameters and disturbances . The control laws based on Sliding Mode Control (SMC), Passivity Based Control (PBC), Model Predictive Control (MPC) and have been implemented on AUTAREP manipulator. A typical response is illustrated in Fig. 4, which shows the ability of a non-linear control technique to handle disturbances.
ROBOTS IN SOLAR ENERGY SECTOR
Robots find enormous potential in production, handling, installation, inspection and maintenance operations in the solar sector. In a typical solar system manufacturing, robots can automate various processes involving silicon mod- ules, silicon ingot, solar cells and silicon wafers. The role of robots in the solar energy sector can be broadly categorized into two domains; (a) Handling cells and wafers and (b) installing, assembling and performing O&M of solar panels.
A. ROBOTS FOR HANDLING CELLS AND WAFERS
Robots are best suited to handle silicon wafers and solar cells owing to their ability to handle delicate components. Compared to the performance achieved with manual proce- dures, robots assemble various solar components in a precise and gentle fashion since they can accurately demonstrate user-defined speeds while ensuring reliability and repeatabil- ity. A prominent example of efficiency improvement in solar cells using the application of advanced technologies is ‘p-type monocrystalline perc solar cell’ by a company Jinkosolar, which proclaims itself as the largest manufacturer of Photo- voltaic (PV) arrays. Marking a new world record, the com- pany improved the efficiency of the cells from 22.78% to 23.45% with the help of technologies like mobile robots, intelligent mobile devices and the Internet of Things (IoT) in the manufacturing chain.
Various robots have been reported in the scientific commu- nity for cell and wafer handling. IRB Flex Picker, by ABB Inc., is an industrial robot used for sorting and handling of silicon wafers and solar cells. It is also applied for loading and unloading of solar cells production units and is considered as the fastest robot in terms of executing pick and place cycles per minute. It offers maximum han- dling of 15g at 200 cycles/min. The cycle time depends upon the tool, gripper, path radius, etc. It is an inverted mounted robot with 120-145Kg weight having a payload capacity up to 8Kg. Another functionally similar robotic manipulator used for pelletizing/depolarizing, handling and loading/unloading applications is IRB 4600. It can handle heavy payloads of up to 20Kg with flexible mounting capability. Another prominent name of the global mechatron- ics solution provider is Stäubli Corporation, which offered 4-axis and 6-axis robotic manipulators to handle all crys- talline (c-Si) production processes. Examples of the robotic arms include ultra-precise 4-axis FAST picker TP80, SCARA (Selective Compliance Assembly Robot Arm) TS and 6-axis TX manipulator.
Cutting of solar cell modules and edge trimming is another important step in the production of solar panels. IRB
6640 series of ABB industrial robots offer high precision to achieve this step in the shortest cycle time. This series of robots provide a reach from 2.55m to 3.2m with a handling capacity from 130Kg to 235Kg. Further, for the preparation of ribbons and soldering of solar cell modules, IRB 1600 series of ABB robotics is an efficient deal. While ensuring quality and reliability, the robot offers 1.2m to 1.45m reach and can handle 6Kg to 10Kg payloads with almost up to half-cycle time than other competitors [92], [93]. Adept Robotics offers a similar solution for solar panel production. Overhead mount robot, Quattro s650, is a parallel robot that offers the largest work envelop and handles solar cells at the highest speed with minimum breakage level [94]. It can handle a payload of up to 6Kg. Adept Inc. further facilitates solar panels’ cutting, sorting, assembly, testing, loading/unloading, and inspection with its Viper and Cobra modules.
In an attempt to strengthen the manufacturing of solar cell arrays with a focus on the space industry, researchers from the Robotic Institute of Shanghai Jiao Tong University have developed a robotic platform to automate various pro- cesses. The platform can auto-dispense and can auto-laydown with a three DOF mechanism (see Fig. 6) to handle the large sizes of arrays with high precision. The system effectively takes care of adhesive thickness, avoids bubbles formation and offers stainless production.
FIGURE 6. CAD model of a robotic system for auto-dispensing and auto-laydown of solar cells for space industry.
PV cells are covered with a protective glass coating, whose capability to generate electricity is constrained if dust accumulates on the modules. A study reported in explored the relationship of dust thickness and solar intensity on the power output of a PV module. The experimental setup consists of a fixed tilt angle of 16◦ . The results in the form of power output corresponding to solar intensities of 400 W/m2 – 700 W/m2 for various dust thickness were presented. It was noticed that degradation in the PV performance decreased with the increase in solar intensity. The reduction in power output was negligible at 700 W/m2 compared with a 25% reduction at 400 W/m2 . Another study investigates the effect of varying tilt angles on the transmittance of plates due to atmospheric dust for around one month. Figure 7 presents the results of this experimental study show- ing eight different degradation curves corresponding to tilt angles from 0-90◦ w.r.t horizontal. The results dictate that fractional degradation in transmittance is a strong function of dust accumulation and tilt angle in addition to the exposure period and climate conditions of the site.
Other recent studies exploring the degradation rate of transmittance as a result of deposition of dust and various contaminations are reported in.
Rain helps in cleaning the panel, provided they are slanted downward. However, in desert regions, sand and dust are accumulated on the surfaces. Cleaning by human labor- ers is not a viable solution due to the remote location of panel facilities and harsh weather. Table 1 presents a com- parative summary of cleaning strategies based on water- wash and robotic mechanisms. Autonomous robots with onboard state-of-the-art sensor technology have the potential to replace the traditional method of manual cleaning of the modules. The choice of a particular solution depends on the application domain, geographical terrain, desired perfor- mance and economic factors. A comprehensive study reported in mentions that automatic cleaning is an optimal choice for panel cleaning on Earth as well as on Mars.
Figure 8 illustrates a robotic cleaner conceptualized and realized under the supervision of the author. It is a custom- designed and indigenously developed wheel-based mobile robot equipped with a roller brush, ducted fan and blower fan.
The cost of the cleaner is as low as 50$. Figure 9 demonstrates the potential of the robot in cleaning PV modules.
Recently, mobile robots are being equipped with a multi- DOF robotic arm for cleaning of PV panels. A prominent example is of ‘Solar Panel Cleaning Robotic Arm’ (SPCRA), which is a 4-DOF mobile manipulator with two prismatic and two revolute joints. A single unit containing a wiper, an air blower and a water sprinkler is housed on the end effector. Experimental trials conducted on 50W solar panels demonstrated efficiency enhancement of 9.1%, which can be further improved to a significant extent using modules of a higher rating. Table 2 presents the results of the energy consumption of SPCRA for a one-time cleaning operation.
Other examples of robots cleaning solar panels include GEKKO, E4, RAYBOT, etc. GEKKO is a mobile robot that is claimed to be four times more efficient compared to manual cleaning. The robot can steep rooftops of up to 45◦ and is teleoperated by a joystick. Unlike GEKKO, another robot E4 performs cleaning operation without water using controlled airflow and microfiber to flick away soil. The gravity ensures that soil is wiped in the downward direction and off-panel rows. The robot has a dedicated onboard solar module to meet the power requirements. Thanks to the World Wide Web, the robot can be controlled anywhere from the globe and weather data is available from onboard sensors. Another autonomous robot based on dry-cleaning has been designed by Miraikikai Inc. Japan. It is a small wheeled robot that is powered with batteries and is equipped with advanced sensory mechanism and rotating brushes to remove sand and dust. RAYBOT is also a dry-cleaning robot presented by Ecovacs- a household robotic innovator. It is essentially a Roomba robot equipped with brushes, vacuums and a blower to wipe off sand. The robot has been successfully tested in China and California before its launch.
TABLE 2. Energy consumption of SPCRA for one-time cleaning.
TABLE 3. Popular solar panel cleaning robots.
ROBOTS FOR INSTALLATION, ASSEMBLING AND MAINTENANCE OF SOLAR PANELS
Robots are now becoming an integral part of the solar panels industry. Due to the complex design of the panels and associ- ated stringent requirements in terms of precision, consistency and delicate nature, robots are a natural choice.
Robotics community has offered numerous solutions to facilitate installation, assembly and O&M of solar pan- els. German companies Kiener Maschinenbau GmbH and PV-Kraftwerker jointly realized a solar plant installation robotic system named MOMO, which provides efficient pro- ductivity and offers reduced risk factors. The system, shown in Fig. 10, offers cost-effective assembly, mainte- nance, cleaning and dismantling of solar plants. It can be operated in difficult terrains and weathers for extended oper- ational periods. It is equipped with a gripper for automatic assembly of PV modules with the capability of numerous cycle repetitions. The robotic system can stand-alone cover 70Km of distance during an assembly operation.
KUKA Systems has offered state-of-the-art production lines for solar panel manufacturers . The pioneer assembly line installed in 2011 in Canada facilitated trimming, framing, testing and packaging of PV panels employing three automated lines with five robots for each line. Such assembly lines take care of all production stages of solar panels of different types and dimensions
. Besides, KUKA offers a whole range of robotized solutions for a variety of solar module manufacturing, cell and wafer handling, lamination, crystalline module simulation, etc.
National Renewable Energy Laboratory of the US, in coop- eration with Spire Corporation, has completed a project for automated production line for solar modules during
2003-2007 named NREL’s Photovoltaic Manufacturing R&D (PVMRD). The project comprising of three phases included the realization of large-area PV arrays (5ft by 12 ft), development of automated production tools for large scale manufacturing and design of solar modules simulator. Robots with various configurations were designed, tested and employed for the project. An operational scenario is sketched in Fig. 12, where a pair of SCARA robots is used to install bus ribbons and diodes. The complete module is processed in three parts using a conveyor belt mechanism so that moderate- sized robots can be used.
Robots are serving the installation and construction of solar PV plants making the process safe, fast and more economical.
FIGURE 14. Comparison of power generated by a static panel and a moving panel.
Solar tracking systems find enormous potential in solar energy applications, as highlighted in. Tracking of the panels is important since the amount of energy har- nessed by a panel heavily depends on its orientation w.r.t. the sun. Also, the tracker helps in uniform distri- bution of solar flux over the surface of the collector. Figure 14 presents a typical comparison of power generation profiles in the case of static and moving panels. A significant difference in power generation capability in both cases is evident particularly in off-peak timings of sun.
Robots can offer a cost-effective and efficient solution for solar tracking along two dimensions. A study reported in presents an optimized design of a solar tracking system based on a 2-DOF robot with a parallel mechanism. The constrained optimization procedure avoids singularities and collisions between links/joints to permit large operational workspace. Based on the universal and spherical joints, the mechanism can move within the angle range 0◦ to 90◦ in elevation and −90◦ to 90◦ in azimuth. Figure 15 presents the experimental results in the form of monthly energy assessment compared to fixed solar panels with tilt inclinations of 0◦ and 30◦ . An overall improvement of 17.2% in energy production is reported using the proposed robotic system compared to the solar panel placed at a fixed inclination of 30◦ . Another automated tracking system centered on a mobile robot (QBotix) is reported in. With an electricity consumption of as low as 30 cents/day, the robot is claimed to have an average ability to adjust five solar panels/min. This idea can be extended to multiple mobile robots, which can also function in a coordinated manner for solar tracking of various arrays. Other studies reporting solar tracking systems include.
Furthermore, inspection and maintenance activities of solar plants are also being rendered by robotized systems. MAINBOT is one such platform designed for large industrial plants. The objective was to develop ground robotic vehicles and climbing robots with diversified sensory and manipulation capabilities to navigate in plants following horizontal and vertical paths respectively. The prototype of the robot (shown in Fig. 16a) has been successfully demonstrated on a cylindrical-parabolic collector type solar plant (Fig. 16b) and central tower plant (Fig. 16c). Results of leakage detection and prevention are reported in. As an alternative to climbing robots, recently Unmanned Aerial Vehicles (UAVs) are being used for inspection of solar plants. One such advancement is reported in, where a UAV and a thermographic sensor are used for inspection of a Concentrated Solar Power (CSP). Field trials were performed at an altitude of 20, 40, 60, 80, 100 and 120 m above ground level with cruising speeds of 5, 7 and 10 m/s. Experimental results proved the feasibility of using UAV to perform real- time inspections for detecting anomalous absorber tunes. Table 4 presents the results in the form of the percentage improvement in UAV deployment compared to the manual inspection. Depending upon the altitude and cruising speed, the improvement range in inspection time using a UAV is reported from 85.6% to 98.0%.
Samwel Kariuki 