This issue was one main result of this part of the evaluation, as recommenda-tions were in line with recommendarecommenda-tions given by the expert but also recom-mendations stated by customers.
result in a acceleration but can also result in a deceleration of theproduct de-velopment process. The creative process of design to cover issues regarding accessibility can decrease the speed of the design process but can also lead to more inclusive products. This issue is relevant especially when there are no suitable solutions to accessibility issues yet.
The second sub-hypothesis3.2expands the evaluation to the practical use of the framework in real product development processes and focuses on the suitability to support designers during product development without hindrance. Product design-ers were able to install and test the software framework in their typical environment to validate if the support is productive and can be included in existing processes without hindrance to typical design tasks. Results show a positive acceptance by designers even throughout different user involvement methods as presented in chapter2.4.
In conclusion, two main issues were identified:
∙ The software is suitable to be included in existing product development pro-cesses independently of the user involvement method applied:
All seven different customer involvement methods were covered by participants of the study as presented in section5.2. Although the acceptance by designers depends on the output of recommendations, it is necessary for better acceptance to extend the amount of recommendations.
∙ An improvement regarding more comprehensive scenario information is advan-tageous:
Regarding a more sophisticated scenario comprehension more background in-formation is needed. This improvement can be made by an extended user man-ual including background information or a further presentation in the software framework.
The last sub-hypothesis adds the perspective of product customers by comparing emerged products created using the system with existing ones. In addition, an expert evaluation was conducted to rate the accessibility compared to recommendations given by the system. The comparison and the expert study identified the following issues:
∙ The use of product interfaces relies on readable and understandable labelling information, so recommendations regarding labels and text information are very important.
Participants have had several problems regarding the identification of single
functions with respect to their icons or characteristics. For instance the “SMS”
button on mobile phones could not be identified correctly in Ireland as it is not commonly used for messages as e.g. in Germany. Icons instead lead to overcome barriers related to language or literacy but the functional meaning of logos can also be ambiguous.
∙ It is not advantageous to concern only one single recommendation, all issues must be solved for a product to be more inclusive, otherwise the modification can also imply new accessibility issues.
This issue was raised regarding the washing machine panels, in which one single recommendation was solved, but 10 new accessibility issues appeared.
∙ If all recommendations are included, product interfaces can be used by a wider group of people.
Most participants were satisfied with the new accessibility and stated the same recommendations as the system.
The evaluation concludes a positive feedback from both perspectives designers and end customers, but with additional comments. The system is capable to be included into real product development processesand does not affect existingproduct devel-opment processes as an obstacle in typical product design. Also a strong learning curve was observed, raising context awareness of end customers on designer side.
However a broad expanse of recommendations regarding product interaction is re-quired helping designer in the creation of suitable inclusive design for an as wide group of end customers as possible.
Discussion and Future Work
6.1 Discussion
The aim of this thesis was a contribution to support product designers during the product development process solving the problem of inclusion of beneficiary needs. More precisely designers should be able to access contextual information about customers of their product to include related issues as early as possible. A framework was implemented and applied in industrial field.
The data used in the system has to be widely extended, so also cognitive impairments can be included by the addition of new classes and rules analogously as existing impairment profiles described in section 3.2.1 and 3.4.4. Regarding the definition of cognitive parameters and the classification of User Models into no, mild and moderate cognitive impairment groups, user trials are necessary. Also different new target products can be implemented.
One main factor during this thesis was the software not restraining the designer.
A possible software framework would add the possibility for designers to create a virtual model of their product, press one big ”Start” button and afterwards the product is inclusive. The software would change the complete design to end user needs itself. On the one side, it would be great to have such a solution, but this would also result in a smaller variety of product designs. Each product would only focus on the connection of technological and human factors by guidelines and conditions. This extreme scenario results in inclusive design, but would also destroy creativity during product development process. The amount of different designs would be narrowed by the target device, resulting in very similar end results.
The presented framework can be seen as a first supportive concept regarding inclu-sive design and was proven that it already can help during product development
processes. But for a seamless integration without restrictions even to devices, a hierarchical superstructure about various kinds of interactions is necessary. As seen in section 1.1, during the last years technology evolved due to new functional possibilities but also new kinds of interaction (speech, gesture etc.). For instance buttons are increasingly replaced by touchscreens both reflecting the same func-tionality. Different devices can provide the same funcfunc-tionality.
It is possible to include a higher stage of hierarchy into the framework presented in this thesis, as different user tests would be necessary to obtain the information, which devices are suitable for which functionality.
When restricted to single devices, results using the presented framework can be optimal. Regarding a more idealistic view, an optimal scenario would be to get rec-ommendations by target functionality (or functionalities).
The target functionality would be the main hierarchical root defining suitable de-vices. The current evolutions such as smart phones are in line with this theory.
Existing devices could be selected by their suitability based on a set of target func-tionalities. Even new devices could be generated by need if a set of functionalities can not be provided by existing ones.