Systems Engineering: RTV Silicone Sealant Application System

agreements Engineering RTV silicone Sealant Application SystemAbstractAs technology advances plainly exponenti exclusivelyy in the 21st century, the need for to a greater extent than(prenominal) and more tangled organisations grows too. Continuous improvement is key to a successful, growing business. This envelops everything within the shaping and engineered placements be no exception. Complex engineered formations require a level of attend, this mark off is important for producing quality products and services. Considering advancing technology and continuous improvement, organisations need to look for ways in which the operation of engineered governing bodys arouse be maximised. Multi-agent bodys (MAS) be a relatively unsanded theory which is put into practice when monolithic establishments cannot resolve the problem, so as trunks become more and more complex, the need for MAS increase. color of TermsFTT defined as the percentage of engines that pass a routi ne first time.JPH the emergence of engines which pass through a process per hour.RTV inhabit temperature vulcanisation silicone sealer.MAS Multi-agent systems1. IntroductionAt the Engine Manufacturing Centre (EMC), cougar Land Rover (JLR) manufacture and assemble diesel and petrol engines. The actor is a process engineer within the diesel assembly hall, whose main job role is to improve any(prenominal) assembly processes that negatively pretend first time through (FTT) or jobs per hour (JPH). Currently, the issue that is guide the largest impact on said deliverables is the automatic application of RTV silicone sealant to the engine block to form a seal with the rear indemnify (Figure 1, below).This alter process has an just FTT of 61% and a value of 59 JPH over the past thirty days. The target JPH value across the entire diesel assembly run along is 68, which means this process is causing an average deficit of nine engines per hour causing considerable damage toward s achieving issue targets. The process itself is performed by two autonomous robots whereby one robot applies the sealant and the other robot holds the engine and moves it along a specified path. in that respect argon two HMIs present, one to program each robot. The robot belongings the engine can be programmed with its position, its movement within the six degrees of granting immunity and its velocity. Whereas the only programmable functions are the start, end and speed and feed of the sealant.2. Systems Engineering conduct Cycle StagesThe role of systems picture is to ensure the success of a system, judged by how well its requirements and development objectives are met, its unconscious process in the field and the space of its mapful operating life. Systems engineering aims to establish a technical memory access that depart aid the operational maintenance and the eventual upgrading of the system. A system life cycle is a term used to encapsulate the organic evolution of a tender system, where it begins with a purpose and grows through development into output signal, operation and lastly, destruction.2.1 Concept DevelopmentWhere on that point is a desire for a clean system, the concept development interpret contains the planning and analysis required to claim the need, the feasibility and the architecture for the new system to best fill the needs of the user.There are four main objectives of the concept development actDecide whether there is a market and need for a technically and economically feasible system.Design and confirm the system requirements after exploring unalike system concepts (see figure 3 below). This stage converts the system which has been derived from the needs analysis, into an engineering oriented view for the concept definition and development. When looking at performance requirements it is important to identify the major functions needed to complete the actions needed. In the gaffe of this example, its structu ral elements should include, power robot, potency movement, take in speed, and apply RTV. To aid with this natural action a systems engineer would use a function category versus functional media diagram (figure 4).Concept selection, agree on its characteristics and plan for the forthcoming stages of engineering, production and operation of the system. It answers the question what are the key characteristics of a system concept that would achieve the about beneficial balance between cap efficacy, operational life, and greet?1.Develop and validate any technological developments required by the new system.2.2 Engineering Development Figure 4 (below) shows three stages of engineering development. Firstly, the sophisticated development stage incorporates two important purposes. One being the acknowledgement and reduction of risks, the second being the development of system specifications. Secondly, the engineering design kind is considerably more detailed than any stages precedin g it. Usually, this stage offers an opportunity for potential customers to get an early look at the product, who can, in turn, provide valuable feedback to the developers. endurely, the integration and evaluation phase is where the new system is installed and subsequently checked to ensure that it meets customer requirements.2.3 Post Development deep down the post-development phase there are two sub-phases the production phase and the operations and support phase. The system is now being produced, for example, for a manufacturing environment. Occasionally there are unexpected issues that arise within the production of the system which requires a systems engineer to solve to prevent disruptions in the production schedule. Once the system is live, system support is critical. Maintenance personnel should be sufficient until more complex problems arise, where they need to call on the experience of systems engineers.3. Function leave Diagram4. Control Architecture4.1 Centralised Contro lThe centralised control system architecture has one component designated as the controller which is accountable for managing the execution of other components. The term architecture is used to suggest a focus on the relationship between the major structural elements in a system. This architecture falls into two classes depending on the execution of the controlled components, any sequentially or in parallel. These are the call-return regulate, only applicable in sequential systems, and the manager model, used in concurrent systems 3.The main movements to use centralised control architecture is that it is simple to conceive and due to its omniscience it can make optimal decisions which take all factors into account. However this architecture does drop drawbacks, most notably the expense in which is required to create the control architecture, the control algorithm needs to be very complex. Furthermore, the degradation of any signal path can cripple the function of the entire sy stem, so they can be fragile.4.2 Hierarchical ControlOrganised in a hierarchical tree, this control system decomposes the problem and allocates it to separate controllers which take control of a subset of the system functions. This can exist over a number of levels, meaning each function could be controlled singlely. Optimal control is liquid possible within a hierarchical architecture as there is always a path to a top-level node however, not all information can travel through every path. Commonly roughly filtering of data occurs between levels.In contrast to centralised control, the control algorithm is much simpler due to decomposition. This means the time and cost of implementation are much lower. Between the diametrical branches of the structure, there is a degree of independence, bring down the effect of system degradation. However, there is usually delay in the affect of each algorithm and in the feedback loop.4.3 Heterarchical ControlHeterarchical control architecture is more robust than hierarchical control and is very flexible and extensible. Additional system functions, such as manufacturing processes and equipment, can be added with almost no added system control cost. However, heterarchical architecture lacks centralised visibility of the system as a whole which means planning can be sub-optimal this control system is sometimes referred to as being short sighted. Though this does mean that short decision making is very good. 44.4 RTV Robot Cell Control SystemCentralised control is not suitable for the RTV robot booth system. The reason for this being the fact that it is too expensive to create and change. Additionally, the fault valuation reserve of the control system must be taken into account. A manufacturing line with such high demand for machine availability must not be crippled by the loss of just one signal.The most suitable and, as it happens, the current control system architecture for this system would be hierarchical. The main d ownside to this architecture is its response time when there are lots and lots of levels. However, its strengths are combines the strengths of the other two control architectures discussed, albeit meagerly diluted. Heterarchical control has strengths that would be fantastic for an automated cell in a manufacturing environment, but its weaknesses deem it unacceptable. If one could combine hierarchical and heterarchical architectures and take apart the myopic nature of heterarchy, it could be a system which improves how automated cells are controlled.5. Multi-Agent SystemsA multi-agent system is a system composed of multiple interacting smart as a whip agents. For problems that are too difficult or even impossible for an individual agent to solve, multi-agent systems can be used. Commonly thought of as being computerised, the agents within a multi-agent system could in like manner be robots, humans, human teams or a combination of humans and robots. There are three different types of agentsPassive agents, agents without goals.Active agents, agents with simple goals.Cognitive agents, agents containing complex calculations.Agents can also be reactive or deliberative, this can be represented by the BDI model (figure 6- below).BDI stands for Belief, Desire, Intention where belief is knowledge of the environment, desire is the need to satisfy an objective and intention is the ability to command action(s).Deliberative agents extend the BDI model to include a symbolic model of the external environment- including data and relationships, memory, the ability to plan and the ability to choose between alternative actions.One could make a case to incorporate multi-agent systems within an automated robot cell at the present time. There is a need for configurability, for example, when a new derivative of engine is introduced and the robot has to be programmed to function differently. The system testament need robustness, so if one agent is lost it does not compromise the whole system. However using a hierarchical architecture provides a sufficient degree of configurability and robustness with less cost and complexity. Multi-agent systems provide driving task allocation rather than pre-planned schedules, for an automated robot cell this is not needed, automation needs efficiency in static conditions 5.6. ConclusionThe automated RTV application robot cell is currently in the operational phase of the system lifecycle, it is in need of improvement however it is not the system that needs improving. The system works as it is meant to, however, it is the incorrect system that is in place which is causing the problems.As automotive technology moves towards electrification and autonomous demeanour there will be a need to include more and more multi-agent systems within the vehicles themselves but also within manufacturing systems. There will be a need for greater flexibility, adaptability, reconfigurability and collaboration. Unfortunately, incorporating a multi-agent system in this instance would not have a positive impact on this system.7. References1 Kossiakoff, Sweet, Seymour, Biemer. (2011). System Life Cycle. In Sage, A. Systems Engineering Principles and Practice. 2nd ed. New island of Jersey John Wiley Sons, Inc.. 77.2 Lecture provided PowerPoint slides.3 Ian Sommerville. (2008). Centralized Control. Available https//ifs.host.cs.st-andrews.ac.uk/Books/SE9/Web/Architecture/ArchPatterns/CentralControl.html. lastly accessed 14/02/17.4 J.M. van de Mortel-Fronczak and J.E. Rooda. (1997). Heterarchical Control Systems for Production Cells. . 1 (1), 213-217.5 Various. (). Multi-agent system. Available https//en.wikipedia.org/wiki/Multi-agent_system. Last accessed 20/02/17.