DARPA-BAA-12-14 iFAB Foundry.pdf

Broad Agency Announcement iFAB Foundry

Tactical Technology Office

DARPA-BAA-12-14

December 5, 2011

• Federal Agency Name – Defense Advanced Research Projects Agency (DARPA), Tactical Technology Office (TTO)

• Funding Opportunity Title – iFAB Foundry

• Announcement Type – Initial announcement

• Funding Opportunity Number – DARPA-BAA-12-14

• Catalog of Federal Domestic Assistance Numbers (CFDA) – Not applicable

• Dates o Posting Date – 12/05/11 o Proposal Due Date – 02/17/12 o BAA Closing Date – 06/4/12

• Total amount of money to be awarded – Up to $50 million

• Anticipated individual awards – Single award is anticipated

• Types of instruments that may be awarded – Procurement contract or other transaction

• Agency contacts

Paul Eremenko, Program Manager, DARPA/TTO LTC Nathan Wiedenman, Deputy Program Manager, DARPA/TTO Christopher Glista, Contracting Officer, DARPA/CMO 3701 North Fairfax Drive Arlington, Virginia 22203 E-mail: DARPA-BAA-12-14@darpa.mil Fax: +1-703-248-8002

TABLE OF CONTENTS

Section I: Funding Opportunity Description …………………………………………4 Section II: Structure of Award ………………………………………………………19 Section III: Proposal Requirements …………………………………………………30 Section IV: Proposal Evaluation Criteria ……………………………………………36 Section V: Eligibility Information & Additional Requirements …………………38 Appendix 1: Philosophical Underpinnings of Adaptive Vehicle Make ………….43 Appendix 2: Depiction of the META-iFAB Integrated Tool Chain ……………….55 Appendix 3: Detailed Description of Current iFAB Efforts ……………………….57 Appendix 4: List of Manufacturing Process and Equipment Models ……………63 Appendix 5: Structure and Semantics of Manufacturing Process Models ……….74 Appendix 6: Open Source License for AVM Software, Models, and Data ………86 Appendix 7: Description of the iFAB/FANG Rock Island Arsenal Facility ……..87 Appendix 8: Sample Bill of Materials for a Notional Infantry Fighting Vehicle ...90

I. FUNDING OPPORTUNITY DESCRIPTION

Background

DARPA’s Adaptive Vehicle Make (AVM) portfolio of programs is aimed at compressing at least five-fold the development timelines for new complex cyber-electro-mechanical systems such as military vehicles.1 Under AVM, DARPA is pursuing the development of novel tools for the design and verification of complex cyber-electro-mechanical systems (META2 /META-II3), along with an infrastructure for the collaborative crowd-sourced development of such systems (vehicleforge.mil4). This design infrastructure promises to yield system designs that are computationally verified to be correct-by-construction.

To complement this design capability, the AVM portfolio also encompasses an effort to develop an information architecture to enable rapidly-configurable, distributed, digital manufacturing. This ongoing manufacturing effort is called iFAB.5 The present solicitation is for the development of an iFAB Foundry, i.e., an actual instantiation of this foundry-style6 manufacturing capability for the domain of military ground vehicles. The iFAB Foundry will provide a distributed manufacturing capability that includes the selection of participating manufacturing processes and equipment, the sequencing of the product flow and production steps, the generation of machine instruction sets, and the generation of instructions for human workers. It is anticipated that the iFAB Foundry capability is likely to result from the amalgamation of existing fabrication capabilities from a model library that characterizes the salient attributes of each modality of fabrication: cost, speed, range of applicability, speed of reconfigurability, etc. The resultant iFAB Foundry need not be manifested as a single facility co-resident under one roof; it can be a virtual aggregation of distributed capabilities, sequenced and tied together into a single resultant product flow.

An overview of the ongoing iFAB efforts can be found in Appendix 3. The iFAB Foundry performer would be expected to rapidly mature and augment these capabilities to develop a manufacturing facility to support the fabrication of a series of products culminating in a heavy, amphibious infantry fighting vehicle (IFV) named FANG. The purpose of the FANG--the Fast, Adaptable, Next-

1 See Appendix 1 for a detailed overview of the portfolio and its philosophical underpinnings.

2 https://www.fbo.gov/spg/ODA/DARPA/CMO/DARPA-BAA-10-21/listing.html 3 https://www.fbo.gov/spg/ODA/DARPA/CMO/DARPA-BAA-10-59/listing.html 4 https://www.fbo.gov/spg/ODA/DARPA/CMO/DARPA-BAA-11-21/listing.html 5 https://www.fbo.gov/spg/ODA/DARPA/CMO/DARPA-BAA-11-20/listing.html 6 So termed in a nod to integrated circuit foundries which are comparably flexible manufacturing facilities.

Generation Ground Vehicle--effort is to exercise the vehicleforge.mil, META, and iFAB Foundry capabilities at scale and in the context of a relevant military system. FANG’s functional requirements will mirror those for the Marine Corps’ Amphibious Combat Vehicle (ACV).

The iFAB Foundry will be exercised through a series of three product builds.

Simultaneously with this solicitation, DARPA is also soliciting proposals for a FANG performer to be selected under DARPA-BAA-12-15,7 who will be responsible for staging a corresponding series of three FANG Challenges--prize-based design competitions for progressively more complex vehicle subsystems--culminating in the design of a full IFV. The FANG Challenges will leverage the META design tools and the vehicleforge.mil collaboration environment to significantly change the design experience and open the aperture for design innovation. The iFAB Foundry performer will be required to support manufacturability queries in the course of the vehicle design process. Once a winning design is selected following each FANG Challenge, the complete technical data package for that design will be forwarded to the iFAB Foundry performer. A largely automated process will then commence to develop a product flow and configure the distributed manufacturing network to support the product build. This would be followed by the automated generation of computer numerically controlled (CNC) instruction sets for each piece of manufacturing equipment, as well as instruction cards for human workers and associated training modules. The iFAB Foundry performer would be responsible for managing the distributed manufacturing network, for maintaining metrology and quality assurance/quality control (QA/QC), as well as for the final assembly of the product at a government-furnished facility in Rock Island Arsenal (see Appendix 7). The FANG performer would concurrently exercise oversight over developmental and integration testing in the course of manufacture to ensure that the design emanating from the META tool suite and vehicleforge.mil community is, in fact, correct-by-construction and undertake remedial design fixes where needed.

The ongoing META program is on track to deliver an integrated capability for:

compositional design synthesis at multiple levels of abstraction; design trade space exploration and metrics assessment with structural and information-based metrics of system complexity; formal semantic integration of models across multiple physical and cyber domains; and probabilistic verification of system correctness with respect to realistic context models using model checking and simulation traces. This capability will be embodied in several end-to-end tool chains ranging from a free, open source set of tools; to a mass-market, web-based, cloud-hosted capability; to a high-end commercial tool suite based around

7 https://www.fbo.gov/spg/ODA/DARPA/CMO/DARPA-BAA-12-15/listing.html existing computer-aided design/product lifecycle management (CAD/PLM) products.8

Also ongoing is the development of vehicleforge.mil, a crowd-sourcing site and capability for the collaborative development of cyber-electro-mechanical systems in a manner similar to forge sites in widespread use today for the development of software by the open source community. The principal technical challenge associated with such a capability is a general representation language that is rich enough to describe a broad range of cyber-electro-mechanical components and system designs, yet formal enough that the system can be “compiled” or verified in some manner when a design change is made to some element or aspect of it.

The META tools provide just such a capability. Additionally, vehicleforge.mil is planned to incorporate a novel reputation-based credentialing architecture for users, component models, and design fragments, as well as a virtual world front end interface to provide an immersive graphical experience for the visualization of the design space.

To the extent that the META tools and the iFAB Foundry information architecture are both model-based design environments--one for vehicles, the other for fabrication facilities--they are both predicated on the existence of a rich set of component models, context (environment) models, and manufacturing process models. To that end, there is an ongoing Component, Context, and Manufacturing Model Library 1 (C2M2L-1) effort,9 which will be followed by C2M2L-2 and C2M2L-3 at appropriate intervals, designed to populate model libraries on a cadence corresponding to each FANG Challenge.

It is important to note that the iFAB Foundry represents a novel kind of manufacturing capability, linked by a common model-based information architecture, but that can be naturally distributed both geographically and across corporate and contractual boundaries. Consequently, iFAB Foundry performers need not have the scale or in-house capital infrastructure necessary to develop and operate a conventional factory capable of building an IFV.

In fact, in the interest of maximally opening the aperture to innovative technical and business approaches, DARPA strongly encourages the participation in this solicitation by non-traditional performers, including small businesses, research institutions, and first-time government contractors, so long as they bring a flexible and enthusiastic team and can access top-notch expertise in digital

8 See Appendix 2 for a detailed overview of the META tool chain capability. A demo of one of the current META tool chains can be found at http://tinyurl.com/meta-x-1 in the form of a short screen capture video. Note that this is a snapshot of capability as it currently stands, pending an additional year-long maturation effort.

9 https://www.fbo.gov/spg/ODA/DARPA/CMO/DARPA-BAA-11-47/listing.html manufacturing, model-based design, and domain knowledge specific to vehicle fabrication processes.

Program Overview

The manufacturing capability demands on the iFAB Foundry are centered around three FANG Challenges. The goal of the challenges is to sequentially apply the AVM tools and processes to products of graduating level of complexity, but aligned toward the ultimate goal of building the FANG IFV. The three FANG Challenges will be as follows (and described in detail in a subsequent section):

1. Mobility/Drivetrain Challenge

2. Chassis/Survivability Challenge

3. Full Vehicle Challenge

Each challenge is intended to be a stand-alone design-build-test exercise. In other words, the products of the first two challenges do not necessarily feed into the Full Vehicle Challenge. Prior practice in open source software development and crowd sourcing suggests, however, that significant design re-use will occur in the vehicleforge.mil community and should be encouraged.

The total period of performance for the iFAB Foundry program, during which the three FANG Challenge builds will be conducted, is 36 months from date of contract award. Proposals should be structured as a 12-month base effort, followed by two sequential 12-month options. The technical effort is described for clarity as seven distinct tasks. However, due to the integrated and very fast-paced nature of the technical effort, proposals should encompass all seven tasks and all three FANG product builds. The tasks are as follows (and described in detail in a subsequent section):

1. Integrate iFAB information architecture & manufacturing model library

2. Establish baseline foundry configuration & business/information network

3. Procure and install equipment at final assembly node

4. Support manufacturability queries from FANG Challenge participants

5. Configure foundry for specific FANG Challenge product design

6. Manufacture FANG Challenge product

7. Gather QA/QC and metrology data on manufactured product

Tasks 1 through 4 constitute the development and preparation of the iFAB Foundry manufacturing capability. Tasks 5 through 7 will be repeated for each of the three FANG Challenges to fabricate corresponding products.

Figure 1: Notional Schedule of Principal iFAB Foundry Program Activities

Detailed Description of FANG Challenges

Mobility/Drivetrain Challenge: DARPA has intentionally avoided a precise specification of the scope of the mobility and drivetrain subsystems of an IFV for fear of precluding particularly innovative designs that challenge subsystem boundaries. The general scope of this challenge encompasses: primary drive power generation (e.g., internal combustion, turbine, hybrid, electric, or other powerplant), auxiliary power generation tied to the primary drive powerplant, energy storage, transmission (likely allowing for dual-mode land and amphibious propulsion), suspension, driveline-to-ground interface (e.g., wheeled, tracked, or otherwise), secondary amphibious propulsion mechanisms (e.g., secondary driveline, clutches, waterjets), steering (both land and water), braking, and associated tribology, hydraulic, pneumatic, electrical, cooling/thermal, vetronics, and embedded/control systems (including software for control and fault management systems). Structural elements unique to the mobility subsystem are also in scope, though general elements of the vehicle frame, chassis, and protection system are instead the subject of the Chassis/Survivability Challenge. The final manufactured version of the winning design will be tested on a dynamometer rig for mobility performance, efficiency, ride quality, thermal, and other attributes that can be reasonably observed in the course of standard dynamometer testing. Consequently, for purpose of requirements development and the ability of META tools to synthesize correct-by-construction mobility/drivetrain designs, component and context models corresponding to all of these test parameters will be developed by the C2M2L-1 performers and made available to the FANG challenge participants. The prize award for the winner of this challenge will be $1 million, awarded directly by DARPA. A single copy of the winning design will be manufactured by the iFAB Foundry.

Chassis/Survivability Challenge: The general scope of this challenge encompasses: hull, chassis, frame/panels or monocoque structure, modular armor panels, mountings/inserts, subsystem volumetric compartment placeholders, crew compartment with crew accommodations such as seats and restraints, subsystem mounting placeholders, hull penetrations for drivetrain/grills/ hatches, hatches, and blow-out panels. The final manufactured version of the winning design will be tested using industry-standard complete coordinate measurement dimensional metrology approaches, fit checks, corrosive and environmental effects, and survivability testing for kinetic penetration and blast effects. Consequently, for purpose of requirements development and the ability of META tools to synthesize correct-by-construction chassis/survivability designs, component and context models corresponding to all of these test parameters will be developed by the C2M2L-2 performers and made available to the FANG challenge participants. The prize award for the winner of this challenge will be $1 million, awarded directly by DARPA. A single copy of the winning design will be manufactured by the iFAB Foundry.

Full Vehicle Challenge: The scope of the full vehicle challenge is a complete IFV design, with the exception of the command, control, communications, intelligence, surveillance, and reconnaissance (C4ISR) suite, the battle management system (BMS), and weapons. All three of these items will be specified prior to the challenge and will be supplied as government-furnished equipment (GFE) to the FANG performer for integration into the final vehicle(s) following manufacture in iFAB. It is important to note that the Full Vehicle Challenge is not predicated on the winning designs to the Mobility/Drivetrain Challenge or Chassis/Survivability Challenge as inputs or constraints.

Additional vehicle features beyond those covered in the preceding two challenges include (but are not limited to): crew accommodations, crew interfaces and vetronics, integrated vehicle management systems, fire suppression system, energy (electrical, hydraulic, pneumatic, etc.) distribution, storage, and management systems, environmental controls and ventilation. The final manufactured version of the winning design will be tested against the full range of Marine Corps ACV program requirements in a complete series of operational test and evaluation activities supported by the FANG performer.

Consequently, for purpose of requirements development and the ability of META tools to synthesize correct-by-construction complete vehicle designs, a full suite of component and context models will be developed by the C2M2L-3 performers and made available to the FANG challenge participants. The prize award for the winner of this challenge will be $2 million, awarded directly by DARPA. Additionally, the winning design, upon fabrication in the iFAB Foundry and test and evaluation by the Marine Corps will be eligible for consideration as the go-forward design for the ACV program. To support DARPA and Marine Corps test and evaluation needs, up to eight copies of the FANG IFV will need to be built by the iFAB Foundry. The iFAB Foundry must also be capable of and sized for supporting subsequent full-rate production at a rate of 2 vehicles per week.

Task 1: Integrate iFAB information architecture & manufacturing model library

This task encompasses two principal efforts: the completion and integration of the model-based manufacturing information architecture begun under the ongoing iFAB effort, and the completion and integration of the manufacturing model library initiated under the ongoing iFAB and C2M2L-1 efforts.

The current iFAB effort is described in detail in Appendix 3; the scope of the iFAB and C2M2L-1 manufacturing model effort is summarized in Appendix 4;

while the presently envisioned manufacturing model structure and semantics is described in Appendix 5. Note that several current META, iFAB, and C2M2L-1 performers are also undertaking efforts at formal semantic integration across domain-specific modeling languages. The output of all of these efforts, as well as the opportunity to provide input (under DARPA’s auspices) to help guide these activities, will be available to the iFAB Foundry performer immediately upon contract award.

The iFAB Foundry information architecture should support the following features:

• A manufacturing process model library which provides a semantically formal characterization of manufacturing process capabilities (e.g., range of applicability, range of reconfigurability, cost, speed, tolerances, etc.)

including specific manufacturing equipment, assembly operations, fixtures, and capabilities of human workers.

• The decomposition of a technical data package for a product design into all feasible component fabrication and assembly approaches spanned by a given manufacturing process model library. A product technical data package includes: a META representation of the design as a composition of component models that each include geometric and complete behavioral characterization of the component with formal semantics; a composed CAD model for the product; a standard PLM data set associated with the product.

• A foundry configuration capability which instantiates all feasible fabrication and assembly approaches for a given product design in alternative foundry configurations factoring in sequencing, equipment availability constraints (i.e., scheduling), manpower requirements, and logistics costs (in terms of time, dollars, and uncertainty) associated with a particular geographic/contractual distribution of fabrication and assembly operations. The tool should enable the calculation of and representation of the trade space of feasible foundry configurations vis-à-vis various metrics such as product transit time, production volume, manpower requirements, product cost, and foundry cost. The procurement of raw materials and off-the-shelf equipment should be included into the foundry configuration and metrics calculation.

• Ability to determine whether a particular product design is manufacturable given the scope of a particular manufacturing process model library. It is desirable to instantiate this capability in three ways:

o Answer in near-real time queries whether a particular product tech data package is manufacturable. In cases where a design is not manufacturable, it is desirable to localize the problem to specific aspects or components and suggest what changes can be made to make the product manufacturable.

o Enable the mapping of the space of manufacturable designs spanned by a particular set of manufacturing process models onto the design trade space.

o Derive design rules from manufacturing process models that can be applied to constrain feasible product designs.

• For a particular foundry configuration, generate CNC instruction sets for all manufacturing equipment in their native format, instructions cards for human workers, and training modules associated with human worker instructions.

• Ability to capture in real time the build status of a product, including the state of all components and manufacturing equipment across the entire iFAB Foundry network.

• Ability to capture “as built” metrology and other static QA/QC properties (e.g., materials properties) for components, assemblies, and the final product in near real time as the build progresses such that the tech data package can be updated and re-verified if needed.

• The iFAB Foundry must implement appropriate information assurance controls and physical security measures to handle third-party proprietary and export-controlled designs, components, and products, but will remain entirely at the unclassified level.

While the iFAB Foundry performer should leverage current developments under other AVM and other research efforts (e.g., DARPA Open Manufacturing10 which is also being conducted with an open program construct and shared intellectual property model), the iFAB Foundry performer will be required to exercise sole end-to-end responsibility for delivering the above capabilities, and proposers should plan accordingly.

The deliverables from this task will include a Preliminary (PDR) and Critical Design Reviews (CDR) of the information architecture, tools, and manufacturing model library in fairly rapid succession to ensure that it is ready for implementation and deployment to support the first of the FANG Challenges.

Note that the challenges are intentionally incremental in scope so that manufacturing capabilities and model libraries can be incrementally built up.

Consequently, Delta-CDRs will be conducted prior to the second and third FANG Challenges.

Task 2: Establish baseline foundry configuration & business/information network

Put simply, the iFAB Foundry is a network of manufacturing equipment that is linked by a flow of data (models, CNC instructions, etc.), a flow of goods (raw materials, off-the-shelf items, manufactured components, assemblies, etc.), and flow of funds and promises (contractual relationships). Successful proposers will need to define each of these in significant specificity in their proposal. This task will allow for minor updates to this design, and will then focus on implementing these information, logistics, and business networks.

In theory, a completely open network of distributed manufacturing capabilities should be able to produce most anything. As a practical matter, it will be necessary to constrain the range of capabilities included in the iFAB Foundry so that the manufacturing model libraries are of tractable scope, the computational burden of the foundry configuration algorithms is reasonable, and the number of business relationships that the iFAB Foundry performer needs to form with manufacturing capability provides (“iFAB nodes”) is manageable. To that end, the iFAB Foundry capability can be constrained to the domain of military ground vehicles. DARPA, with the aid of current iFAB performers, has conducted a preliminary assessment of the scope of manufacturing capabilities that must be included in the model library and iFAB Foundry network; this list is reproduced in Appendix 4.

10 https://www.fbo.gov/spg/ODA/DARPA/CMO/DARPA-BAA-11-54/listing.html

Proposers to this BAA should conduct an independent assessment for the purpose of developing their technical and cost proposal based on their own expertise. Task 2 will afford iFAB Foundry performers a limited opportunity to revisit this baseline foundry configuration, i.e., the scope all manufacturing capabilities accessible to an iFAB Foundry user and their corresponding models, so long as it remains consistent with the scope of the contracted effort.

Proposers to this BAA are also required to identify the specific entities (in-house capabilities, other contractors, government labs, etc.) which will serve as providers of the requisite manufacturing capabilities, as well as the details of the business relationship with those entities (type and structure of contract or agreement, cost structure, service guarantees, and other salient characteristics).

Note that if government entities are utilized, they must be independently contracted by the proposer and may not be assumed to be provided by DARPA as government-furnished capabilities. Finally, proposers should define their logistics approach, including identifying the transportation service providers, as well as the details of the business relationship with those entities (type and structure of contract or agreement, cost structure, service guarantees, and other salient characteristics). This task will afford the iFAB Foundry performers a limited opportunity to revisit these decisions so long as they remain consistent with the scope of the contracted effort.

In particular, the iFAB Foundry performer must make co-location and buy-vs-lease decisions for all of the capabilities they deem requisite for the iFAB Foundry. Building 299 at Rock Island Arsenal (see Appendix 7) will serve as a government-furnished location for the final assembly node of the iFAB Foundry.

The remaining foundry activities may be co-located with final assembly, or may be located elsewhere to exploit existing equipment. Any equipment purchased with iFAB Foundry program funds must be located at the Rock Island facility.

Important siting considerations include equipment procurement costs, logistics costs, guarantees of equipment availability, and latency/responsiveness.

Most digitally-programmable manufacturing equipment will still require some modifications in order to make it network accessible and directly programmable by the iFAB Foundry network. Consequently, this task will involve the development or procurement, deployment, installation, and test of appropriate hardware and software. Deployment can be incremental, as needed to support manufacturing capabilities associated with each of the FANG Challenges.

Proposers should note that some traditional IFV production processes with long lead times or a lack of predictability will not lend themselves to the rapid pace of the FANG challenges. Others, such as additive deposition processes, while not traditionally used for military vehicles may be ideal for the AVM scenarios.

Based on DARPA estimates, it is possible that up to 80% of the components used for the FANG challenge vehicle will be off-the-shelf (OTS) parts ordered from commercial or government sources. Thus ordering, tracking, receiving, and assembly functions will be important considerations for the iFAB Foundry and the responsibility of the iFAB Foundry performer. Proposers should describe their approach to establishing the necessary business relationships, particularly where some customization of OTS parts will likely be required.

In scoping the requisite iFAB Foundry capabilities, proposers should assume that a single copy each of the Drivetrain/Powertrain Challenge and the Chassis/Survivability Challenge products will be manufactured. Eight copies of the Full Vehicle product will be manufactured. However, the iFAB Foundry should be sized and scoped to support full-rate production of 2 vehicles per week.

The deliverables from this task will include Preliminary (PDR) and Critical Design Reviews (CDR) of the iFAB Foundry configuration, network infrastructure, and business relationships. These reviews will be conducted in fairly rapid succession to ensure that it is ready for implementation and deployment to support the first of the FANG Challenges. Note that the challenges are intentionally incremental in scope so that capability can be incrementally brought online. Consequently, Delta-CDRs will be conducted prior to the second and third FANG Challenges.

Task 3: Procure and install equipment at final assembly node

The principal objective of this task is the procurement and installation of equipment at the final assembly node of the iFAB Foundry, to be located at a government-furnished facility in Building 299 at Rock Island Arsenal. The facility is described in detail in Appendix 7. The results of the co-location and make-vs-lease trade studies performed in support of proposers’ proposals and updated in the course of Task 2 execution are key inputs to the equipment that is the subject of this task. DARPA is agnostic on the means by which equipment for the final assembly node is procured--it may be purchased, borrowed, leased, or acquired through some other arrangement as long as the capabilities are present and available to support the three FANG Challenge timelines and the potential for subsequent full-rate production.

The iFAB Foundry performer will need to prepare facilities, then procure, install, and certify all equipment. This includes supporting permitting and approval processes as necessary. This includes procuring and installing tools: machine tools, data networking equipment, and software tools. Securing third-party hardware and software licenses, as applicable, is the performer’s responsibility (all software produced under the AVM effort is open source and certain commercial products have special AVM-use licenses). Once installed, the performer will need to verify proper equipment operation and support the Task 1 effort to update models to represent the “as installed” equipment performance.

The iFAB Foundry performer will be responsible for all maintenance of the final assembly node at Rock Island Arsenal, with the exception of those areas specifically dedicated to the FANG performer. Related tasks include but are not limited to equipment maintenance, calibration, recertification, and updating model-based representations of the equipment.

The iFAB Foundry performer shall continuously maintain a master equipment list which will document any equipment procurements, upgrades, and maintenance. Updates to the master equipment list should be included in monthly technical report and the full document should be delivered with the final reports.

Task 4: Support manufacturability queries from FANG Challenge participants

In Task 1, the capability to determine whether a particular product design is manufacturable given the scope of a particular manufacturing process model library was described. One of the instantiations of this capability is to answer in near-real time queries whether a particular product tech data package is manufacturable, and if it is not, to localize the problem and suggest design changes.

In order to deliver this capability to support FANG Challenge participants, the iFAB Foundry performer should develop and operate an infrastructure whereby a vehicle designer can, probably through vehicleforge.mil, submit a tech data package for a system design (which may be a complete vehicle, a subsystem, an assembly, or just an individual component) and receive from an iFAB Foundry server a manufacturability determination, various metrics associated with the time and cost to manufacture that system, and any other information that can rapidly be derived (particularly if the system is not manufacturable) to help the designer improve the design (e.g., make it manufacturable or improve various metrics).

It is desirable for this capability to be autonomous and near-real time. The capability should be implemented in a scalable manner (e.g., cloud-hosted) to support several thousand queries per day during peak times in the FANG Challenge competition periods. If it proves infeasible or uneconomical to implement this as a near-real time service, batch mode processing of requests may be considered as an option.

The algorithms to support this capability will be reviewed as part of Task 1 deliverables. Usage reports should be delivered as part of monthly status reporting and also via short daily reports during FANG Challenge execution periods.

Task 5: Configure foundry for specific FANG Challenge product design

This task takes a specific product design, e.g., the output of one of the three FANG Challenges, and configures the iFAB Foundry for the manufacture of that specific design. The FANG performer will transmit the complete technical data package for the product to be manufactured to the iFAB Foundry performer. The configuration of the foundry then involves several different types of activities.

First, the information architecture and tools developed in Task 1 are exercised in order to produce a trade space of possible foundry configurations and computes various metrics associated with them. In consultation with DARPA, the iFAB Foundry performer selects a particular foundry configuration, including:

• Which manufacturing nodes are employed

• What off-the-shelf items are ordered and from where

• What specific equipment is used

• Which assembly operations take place where

• How all of these activities are sequenced, taking into account any equipment availability constraints

This can be conceptualized as the “product flow” through the iFAB Foundry.

Second, specific instructions sets are generated. These include:

• CNC instructions for digitally-programmable manufacturing equipment,

• Instructions for human workers and associated training modules

• Instructions for the re-positioning of equipment

• Instructions for fabrication of fixtures

• Orders and special instructions to vendors and raw material suppliers

• Planning orders to logistics/transportation providers and shippers

The principal deliverables from this task are manufacturing plans, which are comprehensive reports summarizing all of the above for each product build.

Task 6: Manufacture FANG Challenge product

In this task, the iFAB Foundry performer actually executes the plans and instruction sets generated in the course of Task 5. Components are procured or manufactured, delivered to the iFAB Foundry final assembly node at Rock Island Arsenal, and final assembly takes place.

Since the cost of raw materials and off-the-shelf components for each product build (in essence the recurring cost) is impossible to estimate with any reasonable accuracy without a specific product design, proposers are not expected to incorporate these costs corresponding to the three FANG Challenge builds into their proposals. These costs will be additional scope added to the iFAB Foundry performer’s contract via contract modification. The iFAB Foundry performer will be expected to utilize the tools developed in Task 1 to be able to rapidly estimate recurring costs of a particular design and prepare in a semi-automated manner the supporting documentation (including bases of estimate, bill of materials, and vendor quotes) for a contract modification proposal corresponding to a given product design. Such a contract modification proposal should be developed and submitted to the government within 7 calendar days of receiving a complete technical data package for a given design from the FANG performer.

During the build periods for the FANG Challenge products, the iFAB Foundry performer will be expected to report daily on the status of the build using the real time management and feedback capability of the information architecture as described in Task 1. Other deliverables from this task will be the completed FANG Challenge products.

Task 7: Gather QA/QC and metrology data on manufactured product

The iFAB Foundry performer is responsible for performing quality assurance (QA) and quality control (QC) at every step of the manufacturing process.

Accurate and exhaustive collection metrology and other quality data on the “as built” or “as delivered” constituent components of the system and on the system as a whole is essential to providing closed-loop validation that the system is likely to conform to the correctness properties that were verified in the course of the model-based design process employing the META tool suite. Consequently, the iFAB Foundry performer is expected to procure sophisticated metrology equipment and although such equipment should be subject to the same co-location trade studies as described in Task 2, it is very likely that large-scale metrology will need to be located at the final assembly node of the iFAB Foundry.

While the iFAB Foundry performer is responsible for the collection metrology and other static QA/QC properties (e.g., materials properties) for components, assemblies, and the final product, the FANG performer has additional responsibility for monitoring the “correctness” of the system, i.e., performing developmental/integration testing in the course of system assembly and integration. Although in theory correctness properties are guaranteed through model-based design synthesis enabled by META tools, as a practical matter--at least in the first several product builds--model mismatches, design incompleteness or inconsistencies, and other issues are bound to surface. It is the FANG performer’s responsibility to identify such issues. However, the iFAB Foundry performer will need to work closely to accommodate the FANG performer’s developmental/integration testing in the course of the assembly process, potentially share metrology data, and prepare to accommodate partial re-builds of parts of the system in response to possible re-designs precipitated by integration issues uncovered by the FANG performer.

The iFAB Foundry performer should also employ the QA/QC and metrology data to update the manufacturing model library described in Task 1 with actual tolerances and defect rates for various manufacturing equipment and processes.

The iFAB Foundry performer will deliver with each manufactured product an updated technical data package that includes “as built” metrology data and QA/QC provenance of the product.

II. STRUCTURE OF AWARD

Award Instrument

DARPA anticipates making a single award under this BAA. The award will be either a procurement contract or, where deemed necessary and where appropriate statutory conditions are met, an other transaction agreement (OTA).11 The procurement contract may either be a cost plus fixed fee (CPFF) instrument (or cost plus zero fee in cases where the performer is a non-profit entity) where the awardee has a Defense Contract Audit Agency (DCAA)-approved cost accounting system or a firm fixed price (FFP) instrument in cases where the awardee does not have an approved accounting system, has a preference for an FFP contract, or where the Contracting Officer deems it appropriate. In cases where an FFP contract is utilized, payments will be conditioned on periodic deliverables such as monthly reports so as not to increase the performance risk borne by the awardee vis-à-vis a CPFF instrument.

The award will be structured as a 12-month base period, with two sequential 12-month options, for a total 36-month period of performance. Proposers should structure their proposals accordingly, and both options must be fully priced in the proposal.

Note that the costs of touch labor and of procuring materials and off-the-shelf items for each of the three FANG Challenge builds cannot be estimated until the winning design for each challenge is selected. Consequently, as described in Task 6 above, the iFAB Foundry performer should anticipate proposing a contract modification to cover materials and off-the-shelf items upon the receipt of the technical data package for each FANG Challenge product build.

Contract Deliverables & Reviews

Proposers should propose an appropriate schedule of deliverables and milestones in their statement of work (SOW) with dates indicated as relative

11 Proposers interested in receiving an OTA are asked to submit proposal responses that accommodate both options. The government will evaluate all proposers’ procurement contract proposals in accordance with the established evaluation criteria. After award selection based on these proposals, the government will evaluate the selected awardees OTA proposal with the intent of selecting the program approach offering the most benefit to the government. The intent behind this evaluation approach is to prevent proposers with greater financial flexibility and resources from reducing the proposed cost to the government by providing a large cost share or extra effort beyond that of a contractor with less financial capability.

For further information on OTAs, see:

http://www.darpa.mil/Opportunities/Contract_Management/Other_Transactions_and_Techn ology_Investment_Agreements.aspx.

values after contract award (ACA). That schedule should be congruent or constitute a superset of the minimal deliverables outlined in Table 1 below.

Written deliverables should generally take the form of reports in Adobe PDF format. Where appropriate, enclosures should include complete diagrams, schematics, data sets, models, algorithms, source code, object code, executable code, documentation, test/use cases, and hardware implementing the capability described in this BAA. Where feasible, a flat-file representation of the enclosed item should be included as an appendix to the PDF report. Draft versions of all deliverables (except hardware) should be supplied 30 calendar days prior to the deliverable due date.

Table 1: Technical Deliverables Task Deliverable (format; timing)

Task 1 – Integrate iFAB information architecture & manufacturing model library

• Information architecture PDR (report; 2 mo.

ACA)

• Information architecture CDR (report; 4 mo.

ACA)

• Information architecture Mobility/Drivetrain Challenge IOC (software, models, documentation; 6 mo. ACA)

• Information architecture Chassis/Mobility Challenge Delta-CDR (report; 10 mo. ACA)

• Information architecture Chassis/Mobility Challenge IOC (software, models, documentation; 12 mo. ACA)

• Information architecture Full Vehicle Challenge Delta-CDR (report; 22 mo. ACA)

• Information architecture Full Vehicle Challenge IOC (software, models, documentation; 24 mo.

ACA)

Task 2 – Establish baseline foundry configuration & business/information network

• Foundry configuration PDR (report; 3 mo.

ACA)

• Foundry configuration CDR (report; 5 mo.

ACA)

• Information architecture Chassis/Mobility Challenge Delta-CDR (report; 11 mo. ACA)

• Information architecture Full Vehicle Challenge Delta-CDR (report; 23 mo. ACA)

Task 3 – Procure and install equipment at final assembly node

• Master equipment list--updated in monthly status reports and delivered as part of final reports

Task 4 – Support manufacturability queries from FANG Challenge participants

• Manufacturability query usage report--updates in monthly status reports

• Manufacturability query usage during FANG Challenge design periods (report; daily during

mo. 4-7, 10-13, 18-24)

Task 5 – Configure foundry for specific FANG Challenge product design

• Mobility/Drivetrain manufacturing plan (report; 6 mo. ACA)

• Chassis/Survivability manufacturing plan (report; 12 mo. ACA)

• Full Vehicle manufacturing plan (report; 24 mo.

ACA)

• Full Vehicle full-rate production manufacturing plan (report; 33 mo. ACA)

Task 6 – Manufacture FANG Challenge product

• Build status update (report; daily during mo. 7- 10, 13-16, 25-31)

• Mobility/Drivetrain product (hardware; 10 mo.

ACA)

• Chassis/Survivability product (hardware; 16

mo. ACA)

• Full Vehicle product (hardware; 31 mo. ACA) Task 7 – Gather QA/QC and metrology data on manufactured product

• Mobility/Drivetrain as-built tech data package (models, data; 10 mo. ACA)

• Chassis/Survivability as-built tech data package (models, data; 16 mo. ACA)

• Full Vehicle as-built tech data package (models, data; 31 mo. ACA)

Final Reports

• Base period final technical and programmatic report and delivery of all program data, software, and items/articles (report, models, data, software, documentation, hardware; 12

mo. ACA)

• Option 1 period final technical and programmatic report and delivery of all program data, software, and items/articles (report, models, data, software, documentation, hardware; 24 mo. ACA)

• Option 2 period final technical and programmatic report and delivery of all program data, software, and items/articles (report, models, data, software, documentation, hardware; 36 mo. ACA)

• Technical manuscript summarizing key technical accomplishments suitable for publication in a peer-reviewed journal (report;

12 mo. ACA)

• Technical manuscript summarizing key technical accomplishments suitable for publication in a peer-reviewed journal (report;

24 mo. ACA)

• Technical manuscript summarizing key technical accomplishments suitable for publication in a peer-reviewed journal (report;

36 mo. ACA)

Additionally, certain periodic deliverables will be expected of the performer.

These are summarized in Table 2 below. All deliverables except Monthly Financial & Hours Reports and where the protection of third-party proprietary or Privacy Act information requires otherwise, will be shared throughout the AVM performer community and will ultimately be publicly released where policy considerations and export controls allow. The Monthly Financial & Hours Reports must include the number of hours worked by contractually-identified key personnel in the preceding month. The key personnel hour amounts of these reports need not be auditable figures and may be informally gathered by the performer’s project manager. No draft versions of monthly or weekly deliverables are required; however, draft versions of the PI meeting presentations will be required one week in advance of the PI meeting.

Table 2: Periodic Deliverables Periodic Reporting Items Means of Delivery

Bi-Monthly Presentations and Demos at AVM PI Meetings

PI meetings at major U.S. metropolitan areas (open to all AVM PIs)

Monthly Technical Report Sharepoint (open to all AVM PIs) and e-mail Monthly Financial & Hours Report Sharepoint (open to government only) and e-mail Weekly Informal Progress Updates Video or teleconference (30-45 mins avg duration) Reports per Milestone/Deliverable Sharepoint (access on case-by-case basis) Hardware Items per Milestone/Deliverables

Rock Island Arsenal, IL or another government-designated site in continental United States

Any equipment or other hardware items procured or developed with iFAB Foundry program funds will become the property of the government and shall be left in place or delivered to Rock Island Arsenal, IL at the conclusion of the contracted effort.

Intellectual Property & Data Handling

DARPA desires Unlimited Rights, as defined in DFARS 252.227-7013 and -7014,12 to all deliverables generated by the iFAB Foundry performer under this effort except clearly-identified, widely-available, commercial software tools, with their commercial availability described and substantiated in the proposal.

Additionally, the iFAB Foundry performer should take affirmative steps for open source promulgation of all software, models, and documentation delivered under this effort. To this end, all software, models, and documentation should be licensed in accordance with the terms of Appendix 6 and incorporate the license text.

The above stipulations do not apply to third-party technical data or software that is handled by the iFAB Foundry performer, e.g., as with tech data packages to third-party designs supplied by the FANG performer. The iFAB Foundry

12 http://www.acq.osd.mil/dpap/dars/dfars/html/current/252227.htm#252.227-7013 performer shall comply with the licensing requirements as set by the owner of the design, except that in all cases the FANG performer is required to ensure that each such license provides at least Government Purpose Rights (GPR), as defined in DFARS 252.227-7013 and -7014 (GPR may be provided without reversion to Unlimited Rights after five years), thereby enabling the iFAB Foundry performer to fabricate the design for DARPA as this constitutes a government purpose.

The iFAB Foundry performer must exercise appropriate measures and controls to protect such third-party proprietary information from inadvertent disclosure while in the performer’s custody. Note that FANG and vehicleforge.mil performers have independent requirements for appropriate information assurance controls.

Although the status of individual vehicle component models and tech data packages vis-à-vis the International Traffic in Arms Regulations (ITAR) is currently under review by the government, it is likely that at least some IFV designs will fall within the scope of 22 CFR § 121, The United States Munitions List.13 The following clause will be included in all procurement contracts, and may be included in Other Transactions as deemed appropriate:

(a) Definition. “Export-controlled items,” as used in this clause, means…