The analysis was based on the operational context, concerned with system usage, and system capabilities such as scalability, manageability, performance, reliability, security plus other observable system qualities, and the developmental context, concerned with aspects of system and application development, such as design, coding, programming languages, development environment/tools, software reuse. The resulting requirements are summarized below.
A key factor is providing the flexibility for new researchers to easily develop and deploy useful software, within the period of a typical short research project, and for this the infrastructure must support a high level of abstraction of robot hardware, and the flexibility to use different languages and platforms.
This makes a distributed infrastructure essential, and drives other requirements, such as a code reuse, so that a researcher does not spend a significant time creating the infrastructure.
A. Operational context requirements:
RQ1: Distributed software environment.
RQ2: Application centric not robot centric – Hardware diversity prevents standardizing onboard robot software; the architecture should standardize software interfaces, so an application can be assembled transparently.
RQ3: Support parallel and distributed processing for any application components.
RQ4: Robot independence – an important consequence of RQ2.
RQ5: Support applications with multiple robots.
RQ6: Preserve system and application integrity.
B. Developmental context requirements:
RQ7: Promote a high level of software reuse.
RQ8: Programming language independent.
RQ9: Platform independent.
No comments:
Post a Comment