Data Collection Techniques

The initial development began with an examination of existing interfaces. Open-ended interviews were held with the clients to get a sense of elements that they wish to retain and functionality they want to add.

In preparation for the prototyping plan, additional meetings were held where the existing requirements specification was reviewed. The primary point of dissatisfaction was a desire for the validation requirements to be more focused on demonstrable user interface tests. The requirements were accordingly updated.

For the discussion of the prototype, interface drawings were coupled with a narrative description of a typical program usage. The clients offered comments on elements that were either unclear to them or which the developers didn't understand sufficiently. Potential scenarios for user testing were discussed, but no definitive decisions were made.

Prototype Description

This interface interface is divided into two primary subinterfaces to support the operator in controlling both the overarching disaster response as well as teloperation of specific vehicles. Certain elements of the interface will remain consistent between the two views and they are illustrated in Figure 1:

Figure 1: Common Interface Elements

Figure 2: Orchestration Interface Elements

The orchestration interface provides a high level overview of the entire response scene. It is used by the operator for three primary purposes:

Figure 3: Manipulation Interface Elements

The manipulation provides teleoperation access to an individual robot. It is used by the operator to perform tasks that autonomous systems are currently incapable of. The exact controls and status monitors are dependent on the specific robot being controlled.

Prototype Technologies

To best meet the client requirements, user interface testing for this project will be conducted with a partial implementation prototype. A variety of technologies will be employed in the development of the prototype:

All of the technologies for this application were chosen for their cross-platform capabilities and ease of use. Rapid prototyping packages such as Visual Basic, and prototyping techniques such as mimicking live video with pre-recorded video will not be used for two primary reasons. First, this interface in a completed form may be needed in the near future and time may not be available to reimplement it with more powerful methods. Second, user testing is specifically to ascertain a user's efficacy in controlling a robot, so using prerecorded video is not feasible.

Figure 4 provides a structural diagram describing the connections between the different technologies and the resultant data paths. The system may potentially employ up to three separate computers. Although it may be possible to instantiate each sub-system on the same machine, computational complexity may prohibit this deployment strategy. Currently, the only platform limited technology is the Image Server which has Windows-specific requirements.

The Player libraries and and QT will run on the majority of POSIX-based operating systems, including the Cygwin environment on Windows. Since a client-specified requirement states that the application must run under Windows, the interface application, ImageServer, and USARSim will be running under Windows during user testing.

Figure 1: System Structure Diagram

Figure 4: System Structure Diagram

Prototype Limitations

This protopype is being developed on an extremely abbreviated timeline. Due to these time constraints elements of the ideal interface will have to be remain unimplemented. The elements that will need to be implemented for user testing are:

Elements that would ideally be a part of this interface, but which won't be implemented for time reasons: