5.2.1 Issues to Resolve
Although the system in its current state is functional, two software development problems were encountered during prototype development, which need to be promptly fixed in order to significantly increase the system performance and reliability. The first one is the incompatibility o f the Fire-I camera with the OpenCV camera capture routines, described in Section 3.3.1.3 (other cameras work well). The second, more serious issue is the stability of the ARTag simple fiducial recognition library, described in Section 4.4.6.2.
Another important bit of functionality missing in the current implementation is the P5 glove finger sensor calibration. Presently, calibration can be performed using a third-party configuration application, as described in Appendix C. This functionality should be incorporated into the application instead, as this would simplify the deployment and allow for re-calibration on the go if the sensors change their properties (for example due to increased temperature).
5.2.2 Suggested Areas o f Concentration
The current implementation of the first response device still has several shortcomings which limit the degree to which the prototype can be meaningfully tested in the target environment. Some of them are indicated in the sections below.
5.2.2.1 Wearability
Provided that external battery packs and appropriate mounting gear (backpack or similar) are available, it would be possible to operate the device in untethered mode. The available operating time would be limited in such setup (40 minutes), and the ergonomics would leave much to be desired. Therefore, in future iterations o f the design the following areas need to be investigated and enhanced:
5.2.2.1.1 Power Consumption and Robustness
The mobile computer used for the development of the prototype, Fujitsu Amilo-D 1840 (Section 3.1.1), while offering good processing capability, was not engineered to provide long battery life, or be used extensively in a wearable configuration in adverse environmental conditions. A more suitable computing platform, such as for example the Thermite TVC described in section 3.1.2, should be considered in future iterations.
5.2.2.1.2 Form Factor
In addition to a more suitable wearable computer, the form factor of the peripherals used in the system should ultimately be improved upon. The current prototype utilizes low cost components which, while adequate for proof-of-concept purposes, would not be suitable for deployment in emergency situations.
5.2.2.1.2.1 HMD
The i-Glasses PC3D head mounted display used for the prototype (Sec 3.2.2) significantly occludes the periphery of vision, thus negatively affecting the user’s situational awareness. Integration with existing first response head-wom gear could pose additional
problems. A more innocuous and less invasive option is discussed in Section 3.2.4, where near eye displays are described. Such displays have been recently made available in the consumer market, alongside more advanced, rugged, and expensive solutions useable in more demanding scenarios.
5.2.2.1.2.2 Fim erbend Sensors
The modified P5 glove (Sec. 3.4.2.1) doesn’t provide much flexibility for integration with first responder’s protective gear (e.g. rubber gloves), and a significant portion o f its footprint is used for electronics which are no longer used after the modification. Provided the finger bending sensor based interaction is determined to be a successful technique through user evaluation study, the next suggested step could be investigating the possibility of embedding simplified fingerbend sensors directly in first responder’s protective gloves. According to rationale presented in Sec 3.4.2.3, 3.4.2.4, three sensors should be sufficient. Providing wireless connectivity would be a logical next step for an input device used in this fashion, simplifying donning of the wearable system by the responder.
5.2.2.1.2.3 Camera
Since the head mounted camera needs to be carefully placed relative to the user’s original line of sight, the exact requirements for the camera hardware can be determined only in conjunction with the information about the specifics of the head-mounted display and the protective head gear (helmet, breathing apparatus) to be used. The Fire-I IEEE1394 camera used currently (Section 3.3.1) provides an adequate starting point in terms o f quality and footprint. However, if an even smaller footprint is required, the Reader should refer to the suggestions contained in Section 3.3.3.1.
5.2.2.2 Reliable Wireless Conectivity
At present, the prototype supports IEEE 802.11 connections between the responder device and the network database. The limitations of this technology are discussed at length in
Section 3.5. Several possible alternatives are indicated, including PacketHop (see Sec. 3.5.2.2.1) and Motorola’s Mesh Enabled Architecture (see Sec. 3.5.2.3.1).
5.2.2.3 Testing and Feedback
All aspects of the prototype design, both in hardware and software, need to be thoroughly tested and evaluated by the target group o f users and experts. The feedback obtained thereby should be the fundamental basis for following iterations o f the design, a process ultimately leading to successful deployment o f the device in real world situations. Since technology advances and a growing knowledge base are gained through hands-on usage, this fact should warrant continuing opportunities for improvement.
5.2.3 Project Scope Expansion
As detailed in Sections 2.2.1 through 2.2.3, the prototype system will ultimately function within a broader context, including infrastructure and tools for authoring, distributing and maintaining the annotation database information. Therefore, when the first response device matures sufficiently in future iterations o f the prototype, it would be adviseable to gradually define, develop, and incorporate that encompassing infrastructure into the project. Modularity and open-endedness should be a high priority in these stages, as technological flexibility will facilitate the evolution of the prototype into a complete solution.