Grant: Discovery Grants Program - Individual
University of Victoria
$36,000.00
Towards AI-Supported Automated Software Design
FY 2024-25
Summary
The way we create software is changing rapidly. Our software systems are now built in part using artificial intelligence (AI) support. AI support is being rapidly embedded into software development tools and social coding platforms. In the next ten years, AI tools and technology will be providing critical cognitive support for specifying and designing software systems, building on existing AI strengths in suggesting code changes. **My research vision is to understand how we can support and automate software design using AI.** AI-supported automation will have the *benefit* of simplifying challenging human tasks:
-- balancing software design trade-offs and avoiding technical debt;
-- leveraging existing design knowledge;
-- correctly understanding requirements;
-- managing software community participation, and
-- producing automated and reliable documentation.
But this AI support comes with difficult *challenges*.
- It will be critically important to write clear specifications, which we know to be difficult for humans.
- There are ethical concerns around *what* AI-supported tools suggest, and who controls the underlying training data generating these suggestions.
- It is difficult to make general purpose AI solutions that work for highly context-specific, legacy software systems.
- There is no clarity around where and when AI-supported tools will fail to help, and therefore require humans to take over.
I plan to tackle these challenges by working toward the following goals. Goal 1: I will explore and engineer tests of the design capability and ethics of AI-supported tools, making clear the extent AI-supported automation can effectively supplement human software design activities. Goal 2: investigate how software designers and developers will leverage this AI support using clear specifications. Goal 3: understand the effects of AI support on how non-expert end-users build software.
My team and I will begin with qualitative problem investigations of design support in AI tools, using surveys and interviews with practitioners on simulated design problems. We will then build AI models, leveraging advances in natural language processing, that can make design-relevant code suggestions. We develop techniques to write effective specifications, using human lab studies and novel experimental instruments. Finally, we work with scientific software developers to develop AI support that changes how these end users approach software design challenges.
My research program will:
- Clearly define the capabilities and limitations of AI-supported tools in software design, and critique the current attitude that these tools are a cure-all for software programming problems;
- Improve design approaches to complex software systems, something critically important as software plays an ever-larger role in the Canadian economy;
- Train highly skilled Canadian HQP in the intersection of machine learning, software design, and requirements specification.
Program
Discovery Grants Program - Individual
Recipient
Location
British Columbia, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
RGPIN-2023-05310
Name nom
Ernst, Neil
Department département
Computer Science
Organization id
7
Institution établissement
University of Victoria
Province en
British Columbia
Province fr
Colombie-Britannique
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2023
Award amount
36000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
802
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Information systems and technology
Area of application fr
Systèmes d'information et technologie de l'information
Research subject code
2705
Research subject group en
Information technology
Research subject group fr
Technologie de l'information
Research subject en
Software and development
Research subject fr
Logiciels et développement
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Towards AI-Supported Automated Software Design
Keywords
software design and architecture; natural language processing; human cognitive support; software engineering; ai ethics; interpretability; software quality attributes; requirements specification
Application summary
The way we create software is changing rapidly. Our software systems are now built in part using artificial intelligence (AI) support. AI support is being rapidly embedded into software development tools and social coding platforms. In the next ten years, AI tools and technology will be providing critical cognitive support for specifying and designing software systems, building on existing AI strengths in suggesting code changes. **My research vision is to understand how we can support and automate software design using AI.** AI-supported automation will have the *benefit* of simplifying challenging human tasks:
-- balancing software design trade-offs and avoiding technical debt;
-- leveraging existing design knowledge;
-- correctly understanding requirements;
-- managing software community participation, and
-- producing automated and reliable documentation.
But this AI support comes with difficult *challenges*.
- It will be critically important to write clear specifications, which we know to be difficult for humans.
- There are ethical concerns around *what* AI-supported tools suggest, and who controls the underlying training data generating these suggestions.
- It is difficult to make general purpose AI solutions that work for highly context-specific, legacy software systems.
- There is no clarity around where and when AI-supported tools will fail to help, and therefore require humans to take over.
I plan to tackle these challenges by working toward the following goals. Goal 1: I will explore and engineer tests of the design capability and ethics of AI-supported tools, making clear the extent AI-supported automation can effectively supplement human software design activities. Goal 2: investigate how software designers and developers will leverage this AI support using clear specifications. Goal 3: understand the effects of AI support on how non-expert end-users build software.
My team and I will begin with qualitative problem investigations of design support in AI tools, using surveys and interviews with practitioners on simulated design problems. We will then build AI models, leveraging advances in natural language processing, that can make design-relevant code suggestions. We develop techniques to write effective specifications, using human lab studies and novel experimental instruments. Finally, we work with scientific software developers to develop AI support that changes how these end users approach software design challenges.
My research program will:
- Clearly define the capabilities and limitations of AI-supported tools in software design, and critique the current attitude that these tools are a cure-all for software programming problems;
- Improve design approaches to complex software systems, something critically important as software plays an ever-larger role in the Canadian economy;
- Train highly skilled Canadian HQP in the intersection of machine learning, software design, and requirements specification.