I-Corps: Market Evaluation for Credible Autocoding (MECA)

Georgia Tech Research Corporation, Atlanta GA

Investigators

Abstract

Automobiles, airplanes, and radio-therapy machines are all examples of safety-critical cyber-physical systems, i.e., systems that bring together software and physical artifacts to form intelligent systems of high functional value to society. These systems are called safety-critical, because unanticipated failures can rapidly lead to the loss of lives. This project is concerned with investigating the needs of industry for the development of future generations of cyber-physical systems, with a specific emphasis on software's impact on overall system safety. By properly identifying industry's needs, the team will be able to establish an effective and relevant safety-critical software development process and formulate recommendations about the products that need developing in support of this process. To do so, the investigators will engage into a series of face-to-face interviews with industry representatives, and the findings of the interviews will be shared with business development specialists. To support the interviews, the team will rely on its experience with building a prototype tool that automatically turns high-level safety-critical software specifications into error-free computer programs. Credible autocoding, a term coined by the project team, is the activity whereby cyber-physical software system specifications are transformed into code, together with the verifiable assurance that the code itself meets those system specifications. The hypothesis of this project is that credible autocoding will contribute to reaching the very high levels of safety these systems must achieve in order to be accepted by civil society; the lower development costs enabled by credible autocoding will also ensure that a broader share of the population will be able to enjoy the benefits of the cyber-physical revolution, as this will allow small companies to compete in a market where currently only big players can afford extreme costs of safety-critical software certification. This study will consist of on-site interviews with several decision-makers working with the markets in which credible autocoding might be a necessity. The outcome of these interviews will be presented in a structured fashion to experts accredited by the NSF I-Corps management at regular intervals. A positive outcome would provide a solid base upon which the I-Corps team can obtain funding and proceed with product development activities meeting industry?s needs. No matter what the outcome, these results will give the team an opportunity to adapt its product definition effort. Such an adaptation may occur at all levels, ranging from product packaging to the basic research options pursued by the team in the area of safety-critical, cyber-physical systems. These activities will be supported by Gene-Auto+, the team's credible autocoding prototype developed under prior funding from NSF.