Visible to the public DARPA C2M2L Program


The trade space exploration and design analysis tools in META and the manufacturability feedback and foundry configuration tools in iFAB both require a comprehensive set of component, context, and manufacturing model libraries (C2M2L) as a foundation. The development of these libraries will enable designers to swiftly consider multiple design iterations and evaluate them in a semi-automated manner via the META and iFAB tool chains. Design configurations that meet the program cost, timing, and performance objectives will emerge and progress so that incrementally higher levels of analysis can be used to select the design that optimally meets the stated requirements.

AVM component models consist of much more than geometric CAD data. They contain data describing cyber-physical behavior, material properties, interface constructs and procurement/manufacturing capability in addition to component geometry. AVM component models have the ability to be composed into complex assemblies such that aggregate behaviors are defined by the component models from which they are composed. In order to facilitate the creation and exploration of broad design spaces, a component model library must contain multiple instances of each component type (i.e. 15 engine variants, 12 transmission variants, 6 drive shafts, etc.) with representations across all relevant physical and cyber domains. The designer can then select appropriate components from specific part-class libraries or even have the META tools consider multiple or all options from a part-class library to automatically form a design tradespace.

AVM is working to instantiate a vision where the performance of a system can be validated with comprehensive modeling and simulation to result in a credible, predictably correct system design. Using META tools, system designs composed of models from the component model library will be virtually tested through simulation using context models that simulate the real world conditions that the system will encounter in operation. For the first FANG Challenge held in early 2013, the context model library spanned three environments (atmospheric, land and aquatic) with nine context categories identified for modeling, representing 33 specific environmental conditions related to IFV requirements. AVM is working to significantly expand the context model library to enable blast, ballistic impact, human factors and transportability analysis in preparation for further FANG design events. The revolutionary benefit to the AVM approach is that a complex system such as an IFV can be conceived and virtually tested in a matter of months as opposed to years using conventional approaches.

The AVM vision also calls for nearly automated assessment and feedback of design manufacturability based on the modeled capabilities of the network of manufacturing facilities within the iFAB Foundry. This Manufacturing Model Library (MML) can be thought of as yet another set of contexts to which designs must be subjected. The MML not only informs manufacturability analysis of designs, but also forms the basis for the iFAB tools that generate human and machine instructions to manufacture the design.

In an effort to enable transition, C2M2L is not only focused on model library development, but also on establishing and promulgating standards to serve as the basis for future coordinated modeling efforts across a community of interest. The existence of formalized and accepted model standards reduces the risk and up-front investment faced by new entrants who are eager to take advantage of AVM capabilities. For the ground vehicle domain component model library currently under development, AVM is establishing both a part-class structure that logically organizes the vast list of ground vehicle elements down to the numbered part level as well as a generalized model standard for each of the part-classes that defines all the required elements. Thus, generation of a specific commercial variant of a component model (an internal combustion engine, for example) would only require alteration of the generalized engine model standard with performance specifications, geometry and variation data unique to the identified engine.