The Multi-modal Interface Design Advisor (MIDA) provides multi-modal interface design support to system designers and developers for a wide variety of platforms. It examines the characteristics of the user, the tasks the user must perform, and the environment in which the resulting interface will be used. It also incorporates task network modeling to estimate operator workload and identify resource conflicts. Based on the information gathered from the designer, MIDA automatically inserts all of the code necessary to execute a model and collect performance measures. MIDA then uses these performance measures in combination with a comprehensive database of multimodal technologies and design guidance to generate interface recommendations. Lastly, MIDA allows designers to re-open a project and modify tasks and user characteristics in order to determine the effects of these modifications on the toolís recommendations.
First, the designer begins by entering information about the users of the proposed interface, the environment where the interface will be used, and any hardware and software limitations. This information includes age, disabilities, domain experience, noise levels, lighting, vibrations, any need for protective gear, monitor size, resolution, etc. If the designer is unsure of this information, default values (data sets) will be stored in MIDAís library and the designer can select an environment type (e.g., an office or warehouse) and obtain the corresponding environment information. Designers are also able to save their own data sets into MIDAís library. It is not necessary for the designer to specify all of this information to get advice from MIDA. However, the more information that the designer can specify, the better the design recommendations MIDA can provide.
Next, MIDA asks the designer to map out the job to be performed; including the tasks the users must perform to accomplish their work and the workflow and timing requirements. MIDA incorporates a library of tasks that the designer can select from to create this map or the designer can import an existing model or a task list from a text file or spreadsheet. MIDA then asks more specific questions about each task. These questions vary for each task and change based on the type of task, stage, and type of information required to perform the task. This information is used by MIDA to provide more specific design guidance.
Then, MIDA opens the simulation software Micro Saint Sharp and generates a task network model. Using the drawing tools of the simulation software, the designer will then lay out the paths (i.e., connect the tasks) in order to show their temporal order (sequence).
Once the designer has finished specifying the task sequences, MIDA executes the model in order to estimate workload and identify potential bottlenecks (times when demands are expected to exceed resources) or conflicts (times when users are expected to perform more than one task at the same time).
In Step 3, MIDA reviews the modelís results and uses logic and information about the users and their environment to determine design recommendations. It alerts the designer to potential bottlenecks and resource conflicts and presents multimodal design options to eliminate or reduce resource conflicts, or recommends a redesign of the task if no solution presents itself. MIDA also points designers to areas and technologies that need to be studied further. Finally, MIDA provides guidance regarding when it is advantageous to offer functionality in more than one modality simultaneously or sequentially (for example, a warning could be presented simultaneously as a blinking light, an auditory tone, and a vibration in a joystick; or these redundant cues could be presented sequentially until a response was elicited). This information is presented in a report that gives suggestions for interface components for each task (and the reasoning behind the suggestions).
The ultimate goal of this tool is to optimize the use of multi-modal technologies within human-computer interfaces to reduce operator workload. This can require some iteration through the steps identified above. Once a designer has gone through their first pass of MIDA and reviewed its design recommendations, he or she can go back in the tool to modify user characteristics, environment information, task information, etc. in order to see how these changes affect MIDAís recommendations.
Some of the key features of MIDA are:
- A comprehensive library of multi-modal technologies, design guidance, task analysis methods, usability testing methods, data collection sheets, micro models, and task lists
- Options for importing data and existing models
- Automatic code generation for the task network model
- Automatic data collection and generation of reports, including operator workload and the suggested modalities and interface technologies to perform each task.
For more information about MIDA, please contact firstname.lastname@example.org.