ASSESS Summit 2015

The Analysis, Simulation and Systems Engineering Software Summit was a watershed event bringing together influential representatives from major engineering corporations, leading industry analysts, software and hardware vendors, and distinguished members of the academic and financial communities. These ambassadors to the summit were among the leading thinkers of our industry, and people of substantial influence with the power to effect change. The summit was underwritten by the non-profit Center for Understanding Change c4uc.org and hosted by the Santa Fe Institute on January 8-9, 2015.

The past five+ years have been clouded by massive economic upheavals – and not just from the recession. The purpose of this Summit is to identify and resolve the issues that confront simulation software and the simulation software industry as we move beyond the recent economic crash and face new and complex challenges. We’re defining the simulation software industry as the ecosystem that creates or uses software to analyze, and simulate complex systems/products.

The need for simulation tools, products, and services has never been higher, yet we face major challenges limiting our ability to capitalize on that need. Some of those challenges are obvious to all, some less so – and there are more that we haven’t even considered yet. The Summit is designed to put a stake in the ground and prioritize our issues, expose hidden issues, and start the conversation that we jointly benefit by discussing. Attending the Summit will be a “who's who” of the CAE and engineering software industry: 40 visionaries, each with the title and position that empowers them to implement their vision.

This is a major international summit (by invitation only) in the spirit of the great European summits in the sciences earlier this century.

The ambassadors put 104 issues up for discussion and ended up with 24 key challenges and after much discussion eight issues stood out, seven of which were presented to the larger audience at COFES

• Design-Centered Workflow
• Ease of Use & Usability
• Pre-CAD Analysis & Optimization
• Impact of Web/Cloud/Mobile
• Knowledge Capture & Reuse
• Ability to Combine Heterogeneous Models in a Systems Approach
• Appropriate Model Fidelity & Role of Unsexy Stuff
• Licensing Models Need to be Revisited

Here are the 24 key issues raised at the Summit

Design Centered Workflow

There is a need to move away from analysis or physics focused simulation workflows to Design Centered Workflows in which the simulation being is being performed to drive design decisions effectively & efficiently. This involves multiple aspects focusing on making simulation usable by non-experts in a reliable way and significantly reducing the time to results.

There is a need to move to performing the right level of analysis as quickly as possible to answer the design or behavior question at hand within the appropriate accuracy. The appropriate accuracy is a function of many factors including: the phase in the design process, the impact of error on the intended application, the type of simulation being run, and the time available. The goal should be that of understanding balancing fidelity and time to achieve the best accuracy within time available.

There is a need to move to performing the right level of analysis as quickly as possible to answer the design or behavior question at hand within the appropriate accuracy. The appropriate accuracy is a function of many factors including: the phase in the design process, the impact of error on the intended application, the type of simulation being run, and the time available. The goal should be that of understanding balancing fidelity and time to achieve the best accuracy within time available.

There is a need to move beyond the current approaches related to simulation process automation to capture and reuse of simulation knowledge. This first involves extracting knowledge about simulation definition and results from simulation data. The knowledge required needs to be conceptual or “abstract” in nature and nopt applied to a specific instance (e.g. re-applying nodal loads is not Knowledge).

There is a need to significantly reduce the expertise required for Systems Engineering and Simulation in order to more effectively meet the growing demand driven by business issues. Growth of Systems Engineering and Simulation usage is limited by the expertise required to run valid models. This is further compounded by the rapidly increasing complexity of the systems, interactions and required simulations.

There is a need to evolve licensing models to be more cost effective both for existing users to dramatically increase usage via DSE and System Engineering for Robust Design and for new users to get started. The licensing models are based on the assumption of a small number of expert users running simulations manually or through a process. The current generally accepted licensing models technological advances for DSE and System Engineering for Robust Design and the shift to a broader range of non-expert users. Software vendors are also trying to protect their revenue streams as any business concern should.

There is a need to invest heavily into new tools and technologies to facilitate new user paradigms and scalability across a more diverse user base. The investment needs to happen before the benefit and growth can be realized. The level of investment required is likely beyond the R&D spend that software vendors have available and alternate funding is probably required.

There is a need to leverage technology and approaches from other industries to achieve broader applicability quickly (e.g. quasi physics used in gaming may be good enough for some situations).

There is a need to leverage technology and approaches from other industries to achieve broader applicability quickly (e.g. quasi physics used in gaming may be good enough for some situations).

There is a need for preconfigured IP as models and/or simulation functions to facilitate broader adoption of simulation and Systems Engineering. This requires embedding multiple integrated technologies and integration interfaces to create Simulation plug n play.

There is a need to better manage uncertainty not related to input uncertainty. This requires fundamental validation of the models, processes, and technologies being used before they can be deployed to a non-expert audience. These requirements include Code Validation, Solution Verification, Model Validation, and Uncertainty quantification.

There is a need to better understand and leverage significant changes occurring in computing architectures and how to enable Systems Engineering and Simulation technologies to rapidly take advantage of these new architectures.

There is a need to leverage analysis and optimization techniques, such as topology optimization, to drive very early design decisions at the conceptual and pre-design commit stages. Traditional CAD often does not start until a design commitment has been made. The goal is to leverage Simulation as a design driver to review concepts, compare options, and get to the design to be finalized rather than working with design data fully featured for manufacturing.

There is a need for extensive customization by the end user companies to create packaged applications that can embed simulation, physics and company knowledge. The range of applications required cannot feasibly be delivered by software vendors and sophisticated customization & configuration capabilities are required.

TThere is a need for a new set of tools for simulation, optimization and design of Additive Manufactured parts. The advances of 3D printing have created the potential to use additive manufacturing for production parts bringing along a need to better understand several new aspects of the printed part such as: material properties as a function of the printing. Lattice and string structure concepts.

There is a need to support a rapidly growing set of custom materials as new manufacturing processes and new engineered materials evolve.

There is a need to motivate the Software Vendors to invest as much as possible to take action on multiple fronts to enable a significant change. What sort of incentive/fear combinations can be used?

There is a need for significantly easier to use tools to enable the next generation of users. Issues related to detailed parameters such as mesh criteria and solver settings need to be removed from the process. The new users need to be able to work with reusable definitions of simulations and models that are not instance specific or model specific.

There is a need for a broad range of purpose built “apps” rather than generic simulation tools to enable the next generation of users. This “apps” need to embed knowledge and enable use of simulation for a specific application by the user without the need to develop a comprehensive understanding of the numerical modeling techniques, models and parameters involved.

There is a need within the new users for a crowd sourcing of simulation to enable Experts to share knowledge with non-experts and to facilitate reuse of knowledge as well as enable those who can to perform the analysis.

There is a need for significantly more computing capability to enable the next generation of users. The smarter software and “apps’ will leverage knowledge and require more compute time. These new users are also expected to perform significantly more simulations as a decision support tool and more complex simulations as part of Design Space Exploration or System based Robust Design.

There is a need improve reliability of Systems Engineering and Simulation technologies, methods, data and processes.

There is a need to significantly improve the usability of the simulation and Systems Engineering Tools to enable more effective usage focused at meeting user objectives. This involves making the software more aware of several aspects of the user objectives and accounting for these so that the user can focus on their task at hand. The usability has to focus on enabling the user to leverage Systems Engineering and Simulation to achieve his goals and make decisions in a timely manner without requiring extensive expertise and probably requires layered access.

There is a need to move toward mobile deployment of simulation on any device and any location. The cloud/Mobile aspects will significantly change how simulation and Systems Engineering are deployed and used and will drive expectations on user experience and licensing models.