CC BY-NC-ND
0Available online at www.sciencedirect.com
ScienceDirect
Procedia CIRP 41 (2016) 632 - 637
www.elsevier.com/looate/procedia
48th CIRP Conference on MANUFACTURING SYSTEMS - CIRP CMS 2015
Engineering apps for advanced industrial engineering
Johannes W. Volkmanna*, Martin Landherra, Dominik Luckea, Marco Saccob, Michael Lickefetta
Engelbert Westkämpera
a'Fraunhofer Institute for Manufacturing Engineering and Automation IPA, 70569 Stuttgart, Germany institute of Industrial Technologies and Automation ITIA, 20133 Milan, Italy
* Corresponding author. Tel.: +49-711-970-1943; fax: +49-711-970-1009. E-mail address: Johannes.Volkmann@ipa.fraunhofer.de
Abstract
Today, manufacturing is being shaped by the paradigm shift from mass production to on demand dictated, personalized, customer-driven and knowledge-based proactive production. Thus, shorter product life cycles, an increased number of product varieties, high performance processes, flexible machines and production systems result in an increased complexity in all factory level domains from product design, process development, factory and production planning to factory operation. To handle this complexity, new knowledge-based methods, technologies and tools to model, simulate, optimize and monitor planned and existing manufacturing systems are required.
This paper presents the challenges, the approach and an overview of the results of the EU-FP7 funded project Apps4aME (GA N° 314156) and provides a concise overview over the Engineering Apps (eApps) approach that the project is based on. The project aims at the comprehensive consideration of ICT-based support of Manufacturing Engineering in all the above mentioned domains, called advanced Manufacturing Engineering (aME). The different life cycles are aligned by the development of a Reference Data Model that provides a detailed overview of all relevant domain-specific and inter-domain interdependencies. This life cycle-oriented model enables an integrated product design, process development, factory planning as well as production planning and factory operation. All stakeholders in these activities are supported by eApps that are conceived, developed and validated with the help of four industrial use cases spanning very diverse industrial branches.
© 2015 PublishedbyElsevierB.V. This isanopen access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of 48th CIRP Conference on MANUFACTURING SYSTEMS - CIRP CMS 2015
Keywords: Digital factory; Engineering apps; Industry 4.0; Factory life cycle
1. Introduction
Today, manufacturing is being shaped by the paradigm shift from mass production to on demand dictated, personalised, customer-driven and knowledge-based proactive production. Thus, shorter product life cycles, an increased number of product varieties, high performance processes and flexible machines and production systems result in an increased complexity in all domains from product design, process development, factory and production planning to factory operation. To handle this complexity, new knowledge-based methods, technologies and tools to model, simulate, optimise and monitor planned and existing manufacturing systems are required. Such new tools should allow changes to be made at early design phases to the product and the corresponding manufacturing processes in all factory
structures (from production network to site, area, segment, production system, cells, machines, sensors / actuators) in order to maximise the system efficiency. These tools must be smooth (smart and fault tolerant) in their interaction with human workers as well as working in an integrated way on different shop floor levels along the whole engineering life cycle. With this contribution, the authors present the promising results in studying and developing new solutions of supporting engineers in the holistic product and factory life cycle using mobile and light weight applications.
2. State of the art and challenges
Several engineering methodologies exist in the different domains, such as the guidelines [1] for product design or references models for process development and factory
2212-8271 © 2015 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of 48th CIRP Conference on MANUFACTURING SYSTEMS - CIRP CMS 2015 doi: 10.1016/j.procir.2015.12.031
planning [2, 3]. They enable a systematic engineering of products, processes and factories. Manufacturing Engineering (ME) addresses all interrelated aspects of products, processes, factory and production life cycles, from design and engineering to factory operation, and facility management up to recycling / disposal and re-use. Manufacturing Engineering deals with the main challenges of aligning these Factory Level Domains from product design, process development, factory and production planning to factory operation.
Having a holistic forward-looking view on ME, a strong consideration of the employment of ICT technologies is mandatory in order for the costs to be reasonable in comparison to the effect [4]. Defined as advanced Manufacturing Engineering (aME), the domains from product design, process development, factory and production planning to factory operation supported by ICT technologies enable the modelling, simulation, optimisation, monitoring and visualisation of products, technical processes and factories. The aME is characterised by the following challenges.
The Factory Level comprises the following core domains: product design, process development, factory and production planning and factory operation, as illustrated in Figure 1. Thereby the challenge is the synchronisation and simultaneous generation of all models within the domains by integrating and using or re-using Manufacturing Engineering knowledge in the early stages of the design, planning and monitoring activities [5, 6]. The major challenge is represented by the serial nature of the product, process and factory decision making along their life cycles. From the product requirements, the product design is generated and stored within a product model. Once defined, the process development generates the process requirements (process plan) and the process model from the design features [7]. The factory and production planning domain consists of the selection of the required resources for the factory model and creation of the manufacturing execution programs. After this stage, all the relevant data and knowledge has to be communicated and used for the factory operation.
The life cycle alignment of all Factory Level Domains induces such a complexity that can hardly be handled without the support of ICT technologies. However, the employment of a single holistic manufacturing engineering system seems not to be feasible, with cost and implementation efforts creating a plethora of implications [8].
Fig. 1. Challenge in Manufacturing Engineering
But only the holistic and comprehensive understanding of the Factory Level Domains with all its interrelations enables the task-oriented modularisation of complex planning and optimization activities. Based on this, the single tasks can be supported by application-based and solution-based ICT tools. The ICT tools addressing "mission critical" activities have to focus on the required data, information and knowledge supporting processes that are essential for designing the products, developing the processes, planning and operating the production and the factories. Therefore application-based ICT tools and systems, fully enriched with knowledge have to be developed to support these domain specific tasks with respect to the interdependencies between the Factory Level Domains. To develop these ICT tools, complementary to the life cycle orientation, there is a need for a proper interoperability facilitated by standardised interfaces and exchange formats.
3. Goals, approach and structure
The approach of the project that is described in this paper aims to support Manufacturing Engineering (ME) stakeholders (e. g. product designers, factory planners, production planners, factory operators, ...) with "Engineering Apps" (called eApps further on) to enable an integrated product design, process development, factory and production planning and factory operation. This integration is flexible and is achieved through the deployment, based on a core Reference Model for the holistic planning and optimisation of products, processes and factories along their aligned life cycles. Engineering Apps can be defined as a high performance piece of software or a digital tool, which supports engineers (i. e. product designer, process planner, factory and production planner, factory operator and shop floor manager) in their daily activities with a specific focus on supporting collaboration [9]. eApps are standard-based and intuitive to use. Additionally they can be characterised by being adaptable, context-aware and, if required, easily cross-linkable. The main purpose of the eApps can be further enriched by integrating manufacturing knowledge. These eApps allow the capturing, modelling, representation and sharing of knowledge in all phases of Manufacturing Engineering. They support the re-use of manufacturing knowledge in early phases and the sharing of the manufacturing knowledge between these activities and cover a complexity range from very simple to fully integrated. They can be deployed (online or offline) on multiple devices like tablets, smart phones, PCs using Cloud systems or High Performance Computer Clusters. An approach to the implementation is shown in Figure 2.
Currently available applications and systems support different phases of all these life cycles mostly in an isolated manner, they do not have the capability to be fully and continuously integrated, based on standardised interfaces, in a heterogeneous factory IT landscape. As a consequence, there are communication walls between product designers, factory planners, production planners and factory operators.
Fig. 2. Apps4aME implementation concept
On the one hand, these eApps can be deployed and used not only on new devices like smart phones or tablets, where workers can use mobile solution for decision making on the shop floor in the production and operation phase. This is one key element to support workers directly and to increase the efficiency in the production planning and factory operation by inherent collaboration enabled at the usage time. On the other hand they can be deployed on conventional devices like PCs or High Performance Computers to reach a maximum impact in the Manufacturing Engineering world. Thereby, the eApps should not only fulfil their purpose in the production phase, but also actively support the propagation of information in the product, process and factory planning phases. In this project the flexible integration of the envisioned eApps is realised through modern integration technologies, such as life cycle oriented platforms or using current cloud approaches.
Secondly, with some of the use cases, the employment of semantic ontologies much closer to the shop floor can be tested. Currently, most companies are hesitant with this complex technology. With the solutions envisioned here, one use case fully addresses enabling engineers without IT experience to handle such advanced approaches, while other use cases try the employment of this technology very close to a running production. This is one example, as other new technologies are also under consideration for tests in shop floor environments within the project.
Last but not least, the project promises a verification of the current hype on eApps and therefore in a small part also for the Industry 4.0 approach. The hype is currently cooling down, with industrial users waiting for someone to prove the applicability and value of the new approaches. Using the broad spectrum of industrial branches within this project, it is one of the first real and graspable statements of the effects that can be achieved in reality. These effects can be verified without creating special laboratory environments or with a singular consideration of specially chosen environments.
All three points are of value not only to industrial endusers, but also to the scientific community. This verification should lead to an adaption and therefore sharpening of the overall strategies in this field, even though this discussion may take time to be properly interpreted even after the results are created. While the results in detail will be published separately, this paper presents an overview and puts it into context, as well as show the further way the project and its partners will have to go to further realise the eApps approach.
4.1. Basic research results
4. Results
The project that implements the given approach houses a wide variety of partners, with its demonstration activities therefore spanning a wide variety of industrial branches. Specifically, there are four demonstrators: one in the automotive industry, one in the manufacturing systems industry, one in the food industry and additionally, one general logistics demonstrator that is industry agnostic with the implementation by a large ICT provider and Fraunhofer IPA as an applied research organization. Because all of the demonstrators make use of the eApps approach, the project is able to provide three basic results on very different scientific levels.
First of all, as the approach states, breaking down the walls between life-cycles or even organizational entities promises a deeper and improved use of the existing knowledge on the shop floor. This usage can be extended to the adjacent engineering parts of any production company. In the real application, the effects of this improvement can currently often only be shown for parts of the business processes, due to technical or effort restrictions. New functions become available, that would've been exorbitantly complex to create without the eApps approach.
While the initial approach description already contains different categories of eApps,, it became very clear that for the approach to be graspable for external partners and to be used in other projects, a clearer and more detailed categorization is required. Based on existing classifications derived from current app marketplaces or from software development processes, a new classification matrix was derived. This enables to sort the eApps after a multitude of criteria. The classification is clustered in the following groups:
• Operational area
• Function
• Complexity
O Computational power O Autonomy O Integration
• eApps platform
O Operating software O Hardware
• End device
Each group contains several items, for example to describe the required computational power. The classification therefore helps having critical information available before introducing new eApps into a running engineering environment full of
legacy systems. This information describes the technical implications (e. g. interfaces and hardware requirements) as well as provides sorting possibilities to help find the eApps with the right function that was created having the implications from the specific user in mind.
In business scenarios, data can come from several sources and can be expressed in a variety of forms. Under these conditions, as the use of the exchanged information is complicated, it is necessary that enterprises adopt a new paradigm of collaboration between the various sources of heterogonous data, thus overcoming the problems deriving from their lack of integration. This challenge has been faced in the context of Apps4aME project, where many efforts have been guided towards the standardization of the managed information, in particular within the manufacturing domain. In this regard, as a means to harmonize data from different sources an approach based on the definition of the Core Factory Data Model (CFDM) is used. It unambiguously represents recurring core concepts and their related relationships through a formalism shared within the industrial community [10]. The main expected advantage of this approach consists in the possibility of aggregating and unifying all the information, thus significantly enhancing the semantic interoperability between different heterogeneous systems under the form of agents, services or applications.
Following the approach of recent researches within the factory domain the CFDM is represented through a set of ontologies by adopting the Semantic Web Technologies, which offer key advantages to the whole Apps4aME project because they enable to:
• Represent a formal semantics,
• efficiently model and manage distributed data,
• ease the interoperability of different applications,
• process data outside the particular environment in which they were created and
• automatically infer new knowledge about the concepts and their relationships, starting from the explicitly asserted facts.
The building of a conceptual model from scratch is an extremely expensive activity if the application domain is complex, as it is the case in the given project. Nowadays a large number of very comprehensive reference models is available, covering a wide range of domains. Reusing them can greatly reduce the costs of a new implementation [11]. This is the reason why, at the stage of the CFDM development, it has been referred to the state-of-the-art technical standards covering different domains, for example the Industry Foundation Classes (IFC) and the Standard for the Exchange of Product model data (STEP). In the belief that their reuse has several advantages, a knowledge reuse framework has been defined and formalized, that aims at identifying relevant data models for the formal conceptualization of the four industrial cases, while using the one of the demonstration scenario as a reference case study [12].
Fig. 3. The overlapping knowledge domains for the four industrial scenarios
The results of the framework application showed that one of the valid starting point for' the conceptualization of the industrial cases is represented by VFDM, which aims at formalizing and integrating the concepts of building, product, process and production resource handled by the digital tools supporting the factory life-cycle phases; thus also enhancing the semantic interoperability of these tools [13, 14]. Moreover, the CFDM implementation has been extended according to the requirements of the four industrial cases, in order to represent detailed information concerning the Food, CRM and Key Performance Indicators domains that are not yet included in VFDM. Also, from the analysis of the four industrial cases knowledge domains, various overlaps have emerged (Figure 3).
Last challenge faced during the stage of the CFDM development is derived from the need to revise some classical data management problems, including efficient storage and query optimization for semantic data. For this reason a study has identified a valid semantic repository, capable to handle and reason on huge amount of Semantic data and capable to realize the analyses of intensive data in real-time. This allows collecting and gathering billions of real-time bytes of data on the organization resources, which are then processed instantaneously to optimize their utilization [15].
4.2. Specific applied research results
The four demonstrators of the project are located in different industrial branches. Additionally, they target very different engineering tasks:
• Project management: high level international project management, connecting several stakeholders spread over the world, gathering and creating knowledge to issue a new level of control
• Quoting and production monitoring: execution and monitoring within the production with very little introduction effort as well as increasing the abilities of the systems through semantic ontologies
• Logistics planning: logistics planning, including systems,
knowledge and information, including several sites
• Product monitoring: in line and interlinked product
monitoring from the production line to the usage at the
client.
Looking at these four very different engineering tasks, it becomes obvious, that the demonstrators not only address different activities, but also address engineering activities on different levels. While some demonstrators mainly go deep into detail for a very specific task (logistics planning), others span multiple life cycles and across very different knowledge domains (project management).
In sum, about 22 eApps are being created, depending on the way groups of them are counted. They also range broadly in complexity of their functions, interconnectivity and the other criteria that were described in the classification of eApps given above. This gives the project the ability to verify the effects of very different types of eApps in real production environments. In the following sections, one example of an eApp is explained in more detail, including a description of the effects and the KPIs that are expected to change.
The example is derived from the last bullet in the list of engineering activities above, targeting product monitoring inside of complex production environments. In the case of the Apps4aME project, the partner is from the food processing industry. Food as a product group implies not only a complexity considering its storage, transport and delivery but also is hardly in the main focus of current production ICT providers. This leads to a lack of specifically targeted solutions with the result that existing solutions are adapted only slightly. One of the eApp derived for this case essentially enables a temperature logging, capable to measure the temperature of the product within the required frame. The temperature interval depends on the product group, e. g. for fresh meat between 0 and 2 °C. Products are especially prone to temperature changes during their transport to the client. If the product is not inside of the temperature interval, the whole lot is considered spoiled and needs to be disposed of. This is not only an ecological but also an economical problem, as the responsibility for the product lies with the supplier until it is unpacked from the storage facility of the client and put into their store. Even if it seems illogical at first, it is the simple reason of the requirement to prove that the product was always inside of the temperature interval when it reached the client facility. As it is only unpacked, sometimes up to one week after delivery, this cannot be done without fully monitoring the temperature with an according timestamp. There are temperature logging devices available, but they are expensive and cannot be connected to the ERP system. An ERP connection is required in order to derive the exact composition of the delivery and therefore deriving the applicable temperature limits. Within the Apps4aME project, an eApp was developed to address this issue.
The eApp shows the upper and lower temperature limit, as well as the temperature curve that was read out from the data logger. The data logger is an iButton device, which provides
the capability to measure up to 8 days of delivery time in configurable intervals. The eApp is fully connected to the ERP and can add its information to the order information already available. Using this solution it is easy to prove, when and where in the delivery process, the temperature limit was exceeded. This enables the food industry to shift the responsibility of spoiled food to the actual party that is responsible. As a handheld device, common Sony mobile phones can be used to read out the data loggers with a USB device, which keeps the full price low. This also helps to show the client the full temperature progress directly at the delivery.
Currently additional and more advanced functionalities are in preparation: Especially helpful can be the integration with the reasoning engine also derived in this project, in order to for example suggest the addition of a temperature data logger to a delivery to a client that was prone to problems in the past.
The comparative price (using street prices) of the system is currently about 1/18th of the price of a temperature logger as they are available on the market. As the mobile device is a standard device available on the market, the complete solution is very applicable and supports the existing processes, whilst streamlining them. Next to the lower price, the identification of clients prone to temperature issues is expected to lead to a significant rise in the customers' acceptance rate and therefore to a significant rise in customer satisfaction. Additionally, lowering the number of spoiled food has an ecological as well as a sizeable economic impact.
Over all the resulting eApps, significant impacts are expected and in parts already shown in very different phases of the engineering processes, depending on the demonstrator. Some examples, sorted according to the four demonstrators:
On its highest level, the first demonstrator "project management" addresses the reaction time, when the project lead is asked for the status of the project. This is supported by significantly decreasing the time that is required to access key data, as the systems automatically aggregate information and derive a status on different levels, depending on the type of user. The second demonstrator "quoting and production monitoring" significantly drops the required effort to introduce an IT based monitoring of the production. It also enables the creation of quoting based on new approaches to similarity matches that make a lot better use of the gathered historical data. This helps to get the accuracy of quotes for clients up and with its shop floor management integration also optimizes the lead times in the production. The "logistics planning" demonstrator drastically reduces the time and the complexity for logistics planning, especially for the data retrieval, when two or more production sites need to be tightly planned together. The fourth demonstrator with the "product monitoring" optimizes the delivery process and reduces delivery losses as shown above and also addresses the complex picking processes, as they are required in the food industry.
Altogether, engineering tasks along the whole chain of life cycle are affected and overall, most impacts are either on time efforts that are required for difficult tasks or on the high costs
that usually come with complex and integrated software solutions.
5. Conclusion and outlook
With the project being only a few months before its end, first conclusions can be drawn for the Engineering Apps approach. In sum, the approach works better than expected, keeping implementation efforts low and the solutions having the impacts that are expected. The integration of other modern approaches like semantic ontologies is very promising and are strongly on the way into production environments. Even with the solutions not yet being finished, it is already very clear to see that the impacts that are expected will be achieved in almost all cases with a significant lessening of the required learning curve of the employed eApps in comparison to current engineering software solutions. An unexpected side effect is that most shop floor personnel liked having eApps on their mobile devices which helped with the acceptance. With the eased integration of other legacy systems, the barriers between the life cycles - or on an organizational level, between departments - become lessened, if not broken down completely. While the quantification of the effects the singular eApps are having when being introduced into the real production environment is a task where work has currently just begun, it is already safe to say that the effects are drastic and the introduction efforts are significantly less than with current large scale engineering software solutions. With the demonstrators spanning very differing industrial branches, the effects of the eApps approach that are shown cannot be attributed to specificities of a single industrial branch but are likely generally applicable. Especially with industrial branches that are not the main focus of current ICT providers, the engineering app approach shows better fitting solutions for considerable lesser prices and will have an impact to widen the scope of the digitalization outside of the already well known big players. Nevertheless, it can be shown, that the big players benefit as well from the engineering app approach by reducing the necessity for unprofessionally created and maintained software "island solutions" and therefore promising more control over the complex software landscape.
With the implementation nearing completion, the eApps are currently being deployed at the industrial partners. During this deployment, further and more advanced functions are being evaluated for their worth together with the according end users. Especially additional functions can now be addressed using the semantic ontology and the data warehouse with the integrated reasoning engine. Using the deployment at the end users, the previsioned KPIs will be measured before and after the introduction. This will lead to a quantification of the positive effects that are already visible. As a longer term outlook, the approach is entering a multitude of projects, both scientific and industrial and is on a good way to become a standard approach to introduce innovative software solutions and to address knowledge challenges in several life cycles.
Acknowledgement
This project has received funding from the European Union's Seventh Framework Programme for research, technological development and demonstration under grant agreement no 314156. The authors are pleased to acknowledge the former research team members of the Digital Factory, Fraunhofer IPA, Stuttgart under the management of Dr. Carmen Constantinescu for the inspiration and the foundation for this project.
References
[1] Verein Deutscher Ingenieure e.V. Methodik zum Entwickeln und Konstruieren technischer Systeme und Produkte. Berlin: Beuth Verlag GmbH; 1993.
[2] Verein Deutscher Ingenieure e.V. Fabrikplanung - Planungsvorgehen (Blatt 1). Berlin: Beuth Verlag GmbH; 2011.
[3] Constantinescu C, Westkämper E. Grid engineering for networked and multi-scale manufacturing, In: Mitsuishi M, Ueda K, Kimura F, editors. Manufacturing Systems and Technologies for the New Frontier. London: Springer; 2008. p. 111-114.
[4] Volkmann JW. Methode zur Konzeption und zur Bestimmung der Wirtschaftlichkeit des Einsatzes von Digital Engineering-Systemen. Stuttgart: Fraunhofer-Verlag; 2014
[5] Constantinescu C, Westkämper E. A reference model for factory engineering and design. In: Huang GQ, Mak KL, Maropoulos PG, editors. 6th CIRP-sponsored International Conference on Digital Enterprise Technology, DET 2009. Proceedings. Berlin: Springer; 2010. p. 1551-1564.
[6] Landherr M, Neumann M, Volkmann JW, Westkämper E, Bauernhansl T. Individuelle Softwareunterstützung für jeden Ingenieur: Advanced Engineering Platform for Production. ZWF Zeitschrift für wirtschaftlichen Fabrikbetrieb 2012;107 (9). p. 628-631
[7] Landherr M. Integrierte Produkt- und Montagekonfiguration für die variantenreiche Serienfertigung. Stuttgart: Fraunhofer-Verlag; 2014.
[8] Volkmann JW, Westkämper E. Cost Model for Digital Engineering Tools. In: Cunha PF. Proceedings of the 46th CIRP Conference on Manufacturing Systems CMS. Amsterdam: Elsevier; 2013. p. 676-681.
[9] Westkämper E. Engineering Apps: Eine Plattform für das Engineering in der Produktionstechnik. wt Werkstattstechnik online 2012;102 (10). p. 718-722.
[10] Tolio T, Sacco M, Terkaj W, Urgo M. Virtual Factory: An Integrated Framework for Manufacturing Systems Design and Analysis, Procedia CIRP 2013;7:25-30.
[11] Efthymiou K, Sipsas K, Mourtzis D,Chryssolouris G. On knowledge reuse for manufacturing systems design and planning: A semantic technology approach, CIRP Journal of Manufacturing Science and Technology 2015;8. p. 1-11.
[12] Modoni GE, Caldarola EG, Terkaj W, Sacco M. The Knowledge Reuse in an Industrial Scenario: A Case Study. In: eKNOW 2015. The Seventh International Conference on Information, Process, and Knowledge Management. Lisbon; 2015. p. 66-71.
[13] Terkaj W, Urgo M. Ontology-based modeling of production systems for design and performance evaluation. In: Proceedings of 12th IEEE International Conference on Industrial Informatics (INDIN). Porto Alegre; 2014. p. 748-753.
[14] Terkaj W, Urgo M. Virtual Factory Data Model to support Performance Evaluation of Production Systems. CEUR Workshop Proceedings 2012;886:44-58.
[15] Modoni GE, Sacco M, Terkaj W. A survey of RDF store solutions. In: Proceedings of 2014 International Conference on Engineering, Technology and Innovation (ICE). Bergamo; 2014. p. 23-25
