ISIS IS LEADING EFFORTS TO REVOLUTIONIZE MILITARY VEHICLE DESIGN
Designing a complex vehicle in one-fifth the time is the revolutionary concept behind the Adaptive Vehicle Make (AVM) research program, a flagship initiative funded by DARPA with heavy involvement by ISIS. This ambitious program seeks to drive innovation in design automation and manufacturing and intends to completely revamp the way the DoD supplies vehicles to the nation's troops.
In the conventional model, a contractor creates a military vehicle from the ground up, from design to manufacturing. DARPA's open AVM approach will allow multiple vendors to compete at multiple levels throughout the acquisition process, saving time and money, as well as creating economic opportunities for more participants.
"This is the kind of innovation that allows separation of design from fabrication, a model that revolutionized the VLSI industry. Right now, the design can only be done by a shop that has the integrated capability to do a complete production run. The idea is to detangle the design houses from manufacturing houses and make the entire process more open, innovative, and competitive," said Sandeep Neema, ISIS senior research scientist and research associate professor of electrical engineering and computer science.
In the AVM project, ISIS is creating tools through model-integrated computing and cyber-physical systems engineering processes and methods. These processes, methods, and tools will allow rapid reconfiguration and analysis of the whole vehicle design. Vehicle components can be combined, added, or modified quickly, creating powerful capabilities for designers.
"We should be able to apply these concepts and tools to a variety of vehicles, though the current focus is on amphibious vehicles," said Neema, who along with Ted Bapty, ISIS senior research scientist and research associate professor of electrical engineering and computer science, led the initial AVM design phase, called "META." They are now leading teams of ISIS research scientists and engineers, creating an open-source, fully integrated toolsuite for the broader AVM project.
Through AVM, DARPA is harnessing creativity and activity of unconventional design and manufacturing entities including small companies and self-organized teams of collaborating experts to yield a more open vehicle design process. Acompetition portion of the AVM program feeds into the ultimate goal: the democratization of vehicle design. THis process is open to anyone, even student groups at the K–12 level. The VehicleFORGE collaboration portal (see page 24) will allow diverse teams to work together without geographic limitations.
The major components of AVM that ISIS is leading or participating in are:
- META, which was the initial design automation portion of AVM. It began with concept exploration and small-scale demonstration of the tools supporting the design process. After the year-long initial phase, ISIS was selected as the main "integrator" for the open-source design tool chain OpenMETA and has begun the process of specifying and implementing capabilities that will be ultimately used in the competition phase of AVM. Bapty and Neema are the PIs for OpenMETA. Collaborating with the lead team from ISIS are groups from Georgia Tech University, MIT, Oregon State University, Palo Alto Research Center, Smart Information Flow Technologies, and Stanford Research Institute.
- The Component, Context, and Manufacturing Model Library (C2M2L, pronounced "camel") portion of AVM is developing computer models that are used as building blocks in the design, verification, and manufacturing of vehicles. ISIS is responsible for the integrity and compliance of the component models to accepted and standardized interfaces. More than twenty undergraduate interns helped with the curation of these components. C2M2L will be conducted in phases beginning with the vehicle drivetrain and mobility subsystems. Neema is the PI for C2M2L.
- The Fast Adaptable Next-Generation Ground (FANG) Vehicle portion of AVM sets the rules for competition and defines how that process will be conducted. FANG is run by Ricardo, a Detroit-based engineering and technical consulting company, with ISIS providing training materials, curating all tools, and operating a help desk as a main subcontractor. Bapty is the PI for Vanderbilt's FANG effort.
- The Model-Based Amphibious Racing Challenge (MBARC) is the educational component of AVM (see page 26). MBARC is a student-led design competition that teaches students about model-based vehicle design, the META process, and how to evaluate the AVM design tools. It has a smaller scope than FANG and is performed by student teams from MIT, University of California at Berkeley, and Vanderbilt. Tom Withrow, assistant professor of the practice of mechanical engineering, is the lead for the Vanderbilt MBARC team.
- VehicleFORGE (see page 24) is a cloud-based collaboration hub where AVM designs are submitted, tested, and validated. It includes the creation of a central website to engage several orders of magnitude more talent than is available in the current industry model. Larry Howard, ISIS senior research scientist, is the PI for VehicleFORGE.
- Foundry, as the name implies, is the portion of AVM that will actually build the vehicle. ISIS played an integral part in the precursor to Foundry, instant Foundry Adaptive through Bits (iFAB), which was responsible for the Foundry design. Gabor Karsai, professor of electrical engineering and computer science and associate director of ISIS, was the iFAB PI.
Computer software, hardware, and scads of physical components must integrate seamlessly to meet DARPA's goals for AVM. "We cut across the spectrum of engineering skills to achieve these goals," said Bapty. "The main idea is to drive innovation in design and manufacturing."
"The potential impact of our research on manufacturing is huge," Neema added. "Being able to create a design rapidly and have it built quickly has immense potential. That's the model of manufacturing that the U.S. needs to evolve toward." The success of AVM could yield better products that can adapt to markets faster, with more products rolling out more efficiently and frequently, he said.
META and the other AVM projects represent the culmination of many previous ISIS projects, from designing tools for space station systems to automobile manufacturing information systems. "Having done that process over and over again, we are well suited to meet the AVM challenge. We have the core set of concepts already in place," said Bapty. "Our technology and people base is broad and deep."