Contents:
Problem Analysis and project focus
Tool set Design
The tools’ final design
User Evaluations
Social, cultural and business context analysis

Process

This page outlines in far more depth the process of designing the Digital Powertools toolset, outlining both the process and design decisions.

The first semester was spent setting up the theoretical background, exploring the context of the project, setting up goals, and doing an analysis of the task commonly used on a computer. This was in preparation for creating a consistent toolset for general computer use as opposed to specializing for specific niche applications. This was also where it was decided to design for a tabletop computer, using devices that relate in size to other devices such as a mouse or the 3d navigator.

Problem analysis and project focus

Currently in most of the tangible interaction designs, there are a number of objects (tools, tokens, containers, etc.). These objects are designed specifically for one application, and often even for one particular purpose, and are in most cases a loosely linked token to reference a digital container, or a modifier.

One of the main issues I recognized is that most tangible interaction designs base their tangible tools on a specific application. This means that in most cases, the meaning of a tool is only meaningful in the context of that application, as meaning flows from that particular context. Some designs already abstract it to the level of activities, meaning that the tools could be used in other context for the same activity.

Applications

Classically, tools and interactions are designed on an application level, gathering functionality and interface paradigms in single applications or developer oriented suites.

Activities

Activities involve the actual meaning given by the context; examples are gathering information, or sending messages.

Actions

These are the basic building-stones of activities and applications, combining these leads to functionality. In the case of classical computing generally only clicking, dragging, moving a mouse, typing or scrolling.

Classically, interfaces are designed from the Application perspective towards the Actions. In this project, the tools were designed from the perspective of Actions towards Activities.

Project Goals

• To create a set of tools that can be introduced not as one massive replacement of all interaction, but as a set that can augment a person’s interaction with his or her computer.
• To create a toolset at such a level of generalization that it can be used in most applications, but is specific enough for the actions to be pleasant to use, intuitive and consistent. This set will then allow interaction with a large portion of tasks on a tabletop computer, possibly augmented by application-specific tools for the different applications.
• o create a toolset that can be used in a multi-user situation, where multiple users interact with the same computer and application at the same time.

 

Design Guidelines

Analysing the context and defining project assumptions, the following design guidelines were defined:

• The devices should encompass as much of the functionality of a general purpose computer as possible, but only insofar as it adds value to the interaction.
• Strive to keep consistently digital objects in the digital realm, and the user should interact through the tools, with these objects, instead of pulling the digital world into the physical and interacting with it there.
• Specialized tools for specialized applications are the strength of Tangible Interfaces, as such the toolset should be capable of being used in conjunction with such strong-specific devices.
• Multi-touch Interactive Surface computers are principally capable and likely to be used by multiple users at once. Therefore the toolset should be compatible with multi-user usage of the computer system.
• Users are willing to learn to master a tool, but devices should be intuitive and pleasant to use, and not require intensive usage to learn the initial interactions.
• The devices should cater to a user’s need to decide for themselves how to use the device, or at least allow some expressiveness in the dynamics of interaction.

The second semester was where the process is less linear. As such, in the next few sections, the different iterations will be described as they might have been logically subdivided, as all the different tools went through the same iterations.

Depending on the tool, some spent more time in certain phases of the process, while others skipped certain parts entirely. In the image below, the chronology of the different phases can be seen for each of the tools.

The process during the second semester was a pretty standard iterative design process, except for the fact that it was for a full toolset, which meant each tool went through their own iterations, although the set as a whole was an important aspect of the design of each separate tool.

The Pointer, Keyboard and Navigator did not go through most of these stages directly (although the Keyboard is important in the Designator track). The Opener was quite developed already due to it being a further iteration of the Final Bachelor Project. The Backtracker is the tool that was developed along a linear path, as it was the easiest to design. The Creator went through a number of additional iterations, as it required several redesigns due to lack of intuitiveness and not being consistent in interaction with the other tools.

In the next sections the process will be divided into two sections. The first will be the Task analysis and the design of the tool set as a coherent set. The second section will be the actual design of the different tools, which will first describe the general process followed, and is then subdivided into sections for each tool separately.

Task analysis

During the project, an analysis was made of the different actions performed during normal computer use. This was done by analysing actions performed by myself and other participants. This allowed the observation of some actions that were too obvious to be noticed as such before.

The observations were grouped at first by application, as this was the order in which they were observed. This also helped to group the activities, and served as a context for the actions. Within these applications, a large number of activities were observed. There was a big overlap between the activities, as a large number of activities were the same across applications.

The activities were then analysed on an action level, of which an example can be seen in the above image. The actions were then gathered to see how many individual actions remained. This led to a surprisingly low number of individual actions, out of which a large number were only slight variations, or only varying in context. The schematic representation of these observations can be found in the image below.

A small modification of the Choreography of Interaction approach was used to further abstract the groupings found in the task analysis, combined with the theoretical background of Image Schemas and their metaphorical extensions.

A surprisingly large number of activities actually used a very small set of Schemas, if looked at from an Image Schema perspective. A lot of computer interactions are the same, they only depend on context and dynamic content, but the activities themselves are very similar.

Tool set Design

As originally decided in the project exploration and literature research, the tools were to be designed as tools that would be physical, but which would perform actions on digital objects. Instead of pulling out digital information and mapping them to a physical form through metaphors, the tools would allow the digital information to remain digital and only be manipulated through physical actions. The digital information would always remain in the digital world, and the tools remain in the physical world, both with their own properties.

As such, some of the actions and functionality observed in the task analysis was simply too dynamic to interact effectively without either making speciality devices, or which would simply be too dynamic to be mapped to a specific physical action (for example actions that required infinite controls, such as “Modify Property”, which depending on the property can be handled in many different ways, be it a colour picker or a size scale, or any of the hundreds or even thousands possibly properties of an object).

Using the conclusions of the previous task analysis, a division was made between actions that could be performed with physical tools as opposed to those which would be too dynamic. Most of the actions however were actions that could be done with physical tools. Using the divisions concluded from the task analyses work was started on the set.

Through a few iterations of different groupings, a set of 7 tools was defined. The first few iterations were based on the division found in the image above. Based on the Choreography of Interaction approach and Image Schemas, different groupings were used (the image below), leading to the final set of 7 tools:

Tool Design

Sketching different ideas and designs was used throughout this process, but aside from that the design process can be divided in a few different iterations.

A LEGO iteration was done to use the input from the user evaluations (See User Evaluations) to define possible designs for the different tools. LEGO Technic was used to make different prototypes from both the Movement Sessions and brainstorming sessions. This allowed for a wide variety of explorations (figure 13).

The next step for most of the tools was a prototyping phase. The User Evaluations done on the LEGO Sessions was used to evaluate the prototypes. Using more advanced materials, such as wood, plastics and foam prototyping, the different models that came out of the LEGO Sessions user evaluations were further developed.

After that, the main interaction with the tools was defined, and the form for the different tools was designed. This was done using a combination of wooden modelling and 3d computer modelling. The wooden modelling made it possible to evaluate the models in a real context, to be able to judge size, shape and ergonomics. The 3d modelling allowed for a wide variety of shapes and sizes, trying out different ideas very fast without having to spend days in the workshop. The tools that were only 3d models were then made physical in wooden models.

The next step was making the tools functional and getting the electronics up and running in order to be able to demonstrate the interaction properly, and to make sure the assumptions made in the designs work in a semi-real environment.

Throughout the latter part of the process, the social and business analysis for chapter 7 was made, as well as the formal definition of the interaction design using a document of requirements.

Designator

For the designator, a number of prototypes were made. The main one can be seen in figure 14, which is divided in two semicircles, where the selection and sorting is done by defining an area between the two halves. The problem with this one is that it is very location based, with very little interactiveness, and no clear way to do sorting. There is also the problem of the two separate halves not having a clear connection to whatever action is being done with other tools, as well as there being no clear way to indicate which pair of tools belonged to each other when there were multiple sets on a table.

The Designator was the tool that was the hardest to design. If the other tools can be described as verb tools (to open, to go back, to copy, to paste), the Designator can be both a verb (to select, to sort), but also as the object (in a grammatical sense) of the actions performed by the other tools. This is why the Designator required many iterations.

The decision was made to go towards a single unit, so that any selection made was defined by the one tool, allowing clear connections to any actions, and causing no problems for identification by the applications. A large number of other models, drawings and 3d models was explored. The final design was shaped to designate something on the table naturally. This combined with a simple adjustment slider that affects the size of the selection was a solution to the problem of defining a dynamic selection size without having two devices, or a rotating system (see figure 14), which was too similar to the movements of the opener.

Creator

For the creator, there were two different actions that came out of the Movement Sessions; one was to squeeze something out of a balloon-like item, the other to move like rubber stamping the table.

The balloon-like movement was explored with the model visible in figure 13. The movement to extrude a new item was considered intuitive, but copying a file, or selecting a new template conflicted with the mental model of there being a digital object inside, which would be squeezed out. Because of the nature of the squeezing movement, it was decided to try a few more iterations of the two.

The choice was then made to instead of going for a mental model where a copy is somewhere inside the tool, an imprint is instead made on a rubber-stamp-like device. This change also improved the adherence to one of the design guidelines, in that it kept the tool as a tool instead of a container of digital information.

Due to the difficulty of finding a solution, a bigger number of design sketches and 3d models were made, using individual movement sessions to try out the intuitiveness and consistency of the mental models. This meant that the creator was one of the last designs to be finished.

Backtracker

In the LEGO stage, only a single model was made, because this tool was the one example design from the previous semester. The model was used to test some of the dynamics of the movement. Different options were continuous or discrete movements. The continuous option was intuitive for longer stretches of time, where exploration was mostly in actual time, while the discrete movements was better for switching between states instead.

Using the 3d modelling tool, a large number of different shapes were tried. Because the symbolic shape of a clock was used for the original design, a number of different shapes were tried to see if other shapes would still be as intuitive, but allow for a clearer difference between discrete and continuous history. After exploring a number of different options, it became clear that the circular motion of a clock was not just intuitive because it was a clock, but because the round shape itself is the only physical shape that is intuitively infinite, representing time.

The final design used the circular shape because it fit the continuous interaction, with the pie shape cut out which fit intuitively with the concrete back and forward of different discrete states.

Opener

Due to the opener being a holdover from a previous project (see the Appendix for an unpublished paper on that process), the interaction and movements for the opener were already designed. The only change from the original design was to remove some of the haptic feedback that wasn’t really working.

However the original design’s form was not very modern. While the assignments had given an intro to drawing and mood design, actual form giving was not something offered in a specific module/assignment. During the first semester however, a custom module was set up with Dr. Joep Frens in order to gain a better understanding of forgiving in a more methodical method. A number of results came from that module, amongst which was one new design for the form of the Opener. This final design combined with its mood boards were then used to inform the design of the other tools.

The opener was the original design, and a benchmark for the form giving of the other designs. The electronics and mechanics were redesigned, to fit in with the new form, but in function and interaction the Opener is very much like the original.

Other Tools

The Navigator, Keyboard and Pen are all three already existing devices, which even in their form fit in with the designed tools. Aside from the integration of the Designator with the Keyboard, there was very little to design about these tools, except for their implementation in the software.

Technical implementation and Demo

For the Demo, all 4 of the newly designed tools were made into a final prototype (figures 21-25). The design was based on the Opener design I had made during a model, applied to the other tools as well. Using a combination of turning wood on the lathe, shaping plastics with a heat gun, laser cutting advanced gears and mechanical contraptions, all four tools were made functional hardware.

Tools

For the Creator, an advanced mechanical construction was made with springs and internal moving parts (figure 19), using multiple prototypes, 3d modelling of movements, and laser cut templates in a drawing application. From a mechanical perspective, the Creator was the most difficult to make. The rotating selector was sensed using an optical encoder and a grayscale gradient, while the two other directions of movement were detected using a specially made switch system. Aside from the outside of the plunger which was turned on a lathe, the Creator was made out of lasercut parts, sanded into shape.

The Opener needed quite an accurate gear system (figure 20) in order to open both sides correctly, but as this mechanism had already been previously designed, for this project it required only to be modified for the current shape of the Opener.

The Designator only needed a slider for hardware, so this was relatively easy, although the difficult shape of the Designator took some time to make in wood.
The Backtracker was turned on a lathe, both the big disk as well as the smaller dial, then hooked up using a mechanical rotary encoder.

Table

Due to the fact that connecting a projector to the system in itself would have taken multiple days, I chose to use the screen of the laptop to display the GUI, but the tools themselves were placed on a homemade table made from a frosted glass pane on a custom wooden frame supporting the camera. The camera is a PSeye camera (used on the Playstation 3), but modified to accept infrared light, and with a custom lens that would focus on the table surface

Software

An advanced demonstration software was designed to simulate an environment in which a lot of different digital object were available that could be modified, scaled, navigated, opening, copied, pasted. The software that runs the computer vision is called ReactiVision, which recognises special patterns on the bottom of the tools, and translates this into a position on the screen.

The tools’ final design

This section describes the final design of the tools, as well as the interaction. The interaction will be described using a method learned from an expert who has worked as an Interaction Designer for a software company designing software for the government. I wanted to use a more formal documentation of interaction, instead of the scenarios usually used.

Creator

The Creator is a tool that is used to create digital objects, hence the name. In a computer, a vast amount of digital objects are created in almost every single application. In a lot of cases this might be a document, an image, or even a bit of text, or they can even be collections of other digital objects. The Creator allows the user to store a flat copy of any digital object defined by an application and create an instance of said copy. The Creator can also be used to select a small set of pre-defined templates for digital objects, and then create instances from said templates.
Actions: Create Digital Object, Select Digital Object (for copy), select Digital Object (from templates)

Backtracker

For most people, a digital object is never really gone, it is simply moved to a harder to reach place (the trash), or in a previous version. Any digital object on a computer has a currents state, but in a lot of cases it also has a whole inventory of previous states, either stored in a global undo-queue, or in a local history. The backtracker makes it possible to view and retrieve these previous states of the system as well as of individual digital objects. By positioning the Backtracker on an empty section of the table, and turning the disc in a counter-clockwise fashion, previously deleted objects show up, which can then be reclaimed by dragging them back into the applications. By positioning the Backtracker’s focus point over an existing digital object, the disc can be used to explore previous states of that particular object. By positioning the Backtracker inside an application area (or window), the Backtracker function as an application wide undo/redo queue. Lastly, when a digital object is dragged into the Backtracker’s focus, the object is sent into the Backtracker’s history, performing the task normally associated with the trash bin.
Actions: Explore timeline of Digital Object, undo/redo, remove objects (from current state)

Designator

While performing actions on a single digital object is reasonably possible just based on location, performing actions on set of objects is much harder. Where classical Computer interaction depends on a linear and chronological connection between selection and action, in a multi-user context this is impossible, as other users might perform different action between selection and action. For example if multiple people are typing on keyboards, the computer would not know where to put which text. For this purpose, the Designator was created. The Designator selects a digital object, or a set of digital objects, and can then be either locationally linked (perform actions near the Designator) or relationally linked, by making a virtual connection between different tools and the Designator. The Designator’s selection can be adjusted in size by using the slider on the top, or by adjusting the digitally added modifiers. These modifiers can include sorting (size, date, alphabet and any other custom designed filtering), or selecting application specific presets. For specific situations, a digital version of the Designator can also be used, especially if a selection moves by something other than user action, such as the input cursor for text input.
Actions: Designate target, make grouping, sort content, make links

Opener

The Opener is a simpler tool, in that it is meant only for two actions, opening and closing. These actions however are involved in a significant amount of functionality, as storing information in accessible containers is one of the main purposes of a computer. This means that it depends entirely on the context of either location or what tool the Opener is linked to, what the Opener is actually opening. The most common container is of course the File System, where using a Folder metaphor, information is stored in recursive containers. Another functionality can be to select more advanced templates for the Creator tool by opening its template container. When no application is open on the Computer, opening the tool will result in opening the basic File System, as such giving easy access to either files or applications.
Actions: Open container, Close container

Navigator

The Navigator is an existing tool, being very well-suited for navigating 2d and 3d virtual spaces. It uses a knob that can be moved in 6 axis, and registers them to translate into movements in the 3d space. Depending on the application, a virtual designator can be used to designate the target of the transformation or navigation, or the location of the tool can be used instead.
Actions: Navigate view, manipulate position.

Keyboard

The keyboard is one of the most efficient ways of entering textual data into a computer, and as such should not be replaced by virtual replacements where possible. For a multi-user environment, the Keyboard can be combined with a Designator to allow multiple Keyboards to be used to input text to different containers on the computer.
Actions: Enter data, active functionality (Shortcuts)

Pointer

The Pointer depends entirely on the preferences of the users, but a simple device will always be necessary to select and manipulate the infinite amount of functionality a computer can project on a digital screen. Either a classical mouse, pen input or a touchscreen are possible, depending on the computer’s requirements and the users’ preferences.
Actions: Select target, move objects, activate functionality, modify parameters, create content (drawing).

User Evaluations

For this kind of project, there are two types of user input that can be valuable. One is a qualitative input, where users evaluate and contribute to the project using their imagination and movements to inspire designs. The second is to use them to validate assumptions and design decisions.
Quantitative evaluations are a lot more difficult, since gaining valid results requires the testing context to be realistic, which is almost impossible to do with uniquely new contexts such as tangible interaction on a tabletop computer.
In the FMP, user input has been integrated throughout the project. Four main sessions were used for various input. The next sections will describe the goal for the different sessions in more detail, listing the type of participants, the protocol for the sessions, and the results and conclusions from each.
 

Task Analysis Sessions

At the start of the project, a list needed to be made of the most commonly actions used in day-to-day computer use. So the first user input was simple observation of computer use in order to analyse which tasks were used in day-to-day computer use.

Conclusion

This user input was at the core of the project, and resulted in some of the most fundamental understanding required to complete the toolset. While it might have been possible to simply analyse one person’s usage, looking at multiple and diverse users allowed the analysis to be comprehensive.

Movement sessions

Using a toned down version of the methods developed by Sietske Klooster called Choreography of Interaction [8], movements involving different actions and functionality was explored. Toning it down simply involved restricting the movements to object interaction at a table instead of the full-body movements generally used in the method. A similar approach is the Interaction Relabeling method (Djajadiningrat et al.[16]). The goal of this session was to inspire different movements that were intuitive for actions, outside of the more classical interaction schemes such as shortcuts or menus.

Conclusion

These movement sessions provided all the initial ideas for the different tools, and because they were based on movements instead of functionality, it set the process on track for physicality and tangible objects instead of tokens.

LEGO Sessions

Using the list of tools, initial designs were made using LEGO to make moving prototypes using the manifold tools available in classical LEGO technic. An advantage to using LEGO for these prototypes is that they are very fast to build quite advanced articulated devices, and can easily be iterated. It also makes clear to participants that the prototypes are in progress, and instils a certain playfulness that is good for staying away from existing interaction.

Conclusion

The LEGO sessions were meant as inspiration and evaluation while still in the middle of the design, the LEGO allowed evaluation of movements, without having to design the final shape and interaction, as the LEGO was obviously an intermediate prototype.

Wooden Model Session

As the second to last iteration of the prototypes for this project, the wooden form models were an important milestone. This meant that the intuitiveness and consistency needed to be evaluated with other users. So this final evaluation session was slightly more extensive and formal than the others, which were primarily for inspiration.

Conclusion:

These evaluations showed that the final designs were both intuitive and consistent, even though the tools were not particularly designed to be used without introduction. The intuitiveness was meant to be when in use, but the evaluation showed even without explanation some of the tools were used correctly.

Social, cultural and business context analysis

Researchers in tangible interaction have always had trouble getting any generalized system into commercial use. This is mostly because there is a huge momentum in the personal computer market that makes it really difficult to make changes that affect the whole eco system beyond merely iterative changes. In order for a new interactive system to be adopted, or even tested within a real everyday context, a huge amount of stakeholders would need to be convinced.

Over the past few years, new business processes, as well as societal changes thanks to the internet, have offered a new opportunity for a new kind process to get a technological product into the market without having to convince the powerhouses.

Classical Business Models

Traditionally, in order to get new interfaces into the personal computer market there have been three options.

Standalone augmentation

One business model is in order to design a tangible interface that works purely as an additional way to interact (green line in the image below), which might add value to the interface, but which doesn’t completely replace the existing interactions, just augments. An example of this is the drawing tablet, which was sold purely as a way to make drawing easier, because it was mapped absolutely to the screen, instead of relatively like a mouse or trackball.

3rd party integration

The second business model is to design a tangible interface that is incorporated in a partner’s hardware (red line in the image below). An example of this is the fingerprint readers in some laptops. The fingerprint reader is an added value that allows the company to sell more laptops, while not requiring a big investment to change existing hardware.

Total replacement

The last business model is to completely redesign the interface, this is often done by companies who are completely in control of hardware and software (blue line in the image below). Examples of this was the Palm handheld devices, the Nintendo Wii, some of the Mac pc’s over the years and the original Microsoft Surface project. These products were all complete departures from their competitors, using a specific new interface to go away from more trivial selling points such as performance or storage capacity.

Historical and Cultural phenomenon

Open Source movement

From a software perspective, a very popular concept for a long time has been open source software, which has always had a business model that is now becoming very popular in other areas, where in most cases, profit is made not through the actual product or software, but through ancillary services such as tech support, customization and corporate level integration or through the more recent freemium model.

Open Source hardware

A more recent development has been an increase in popularity of the application of the open source model to open source hardware. Examples of this are the Arduino, the Raspberry Pi and now the Motorola phone. These products have partly or completely published their hardware schematics.

DIY on the internet

The diy generation has really gained some momentum from the internet, with lots of people sharing their own ideas with others, influencing others with their ideas. Of course it started earlier with diy tv shows, but those were mostly intended as hobbies. The tv shows only allowed one expert to broadcast their knowledge to an audience, with the internet, every person could broadcast their own expertise to everyone else.

One of the first big websites that embraced this concept was the instructables website, and later YouTube became very popular, and it allowed all those people to share their expertise through video productions, some even rivalling those earlier tv programs. Home-made content creation on YouTube has also really gained traction, with a huge amount of people earning respectable incomes, and a smaller elite earning huge amounts, comparable to complete tv productions.

Homebrew hardware

More recently, the diy generation has been gaining ground even in a field which used to be reserved for industrial businesses, the creation of actual complex physical products.Some people used putty, woodwork or other methods of making simple products. Starting with DIY CNC routers, then 3d printing became cheaper and two different movements became relevant for this project.

One is the homemade 3D printer movement, the second movement is a growing community around Shapeways [L10] and its ilk, a company that offers to 3d print community designs on commercial and industrial 3d printers, making more complex processes available to diy enthusiasts, such as metal printing, laser sintering and ceramics.

As a submovement, 3D Hubs, has taken the example of shapeways, and make that even more crowd-based. Where Shapeways owns and operates most, if not all, 3d printers, 3D Hubs allows people with their own 3d printers to provide their services to local people through a unified interface on the 3D Hubs website, while people can pick up their 3d prints cheaper and faster at local 3d printers.

Crowd-Patronage

There is an incredible strength and assertion of the new generation of people growing up where every thought is shared and personal entitlement is high. People that do not want to feel like they are being told what they want by big corporations. A new incarnation of an old concept, patronage, has come up where the consumer is given a certain amount of direct power over new product development through businesses like Kickstarter and Indiegogo. These are websites where people can pitch their products not to big money investors or other companies, but to the consumers directly. This crowd-patronage allows people to vote with their money, choosing to fund projects that they really want to try.

The opportunities for the Digital Powertools

So in general there are three movements now:

1. There is the growing tendency for people to create or modify existing products in novel ways to either add new functionality, alter functionality to fit different situations, or to personalize devices in aesthetic or functional ways.>
2. Alternatively there is the opportunity for individuals without financial backing to create quite complex products and systems through services like Shapeways, Sparkfun and Etzy.
3. And finally there is the platforms of kickstarter and indiegogo that allow projects to be funded on a much larger scale, but all directly from the people that want to support and fund those projects because they would personally want them to be made, not because investors in a big corporation believe it will make them money.

The great thing about these new developments is that a very large market has been created for people to try, buy or create completely new products that would normally require either a massive commitment, or for some large company to switch to a new system. People are now free to choose to make their own decisions on how they wish to interact with their computing devices. A massive new channel has appeared for smaller products, companies and projects to reach a small but widespread market (see figure below).

It does have some other effects on the design of the projects though. Especially designing for smaller volumes is a different. The designs need to be designed not for mass production, but for one-off or small batch production from a purely technical point of view. But because the people who are involved in the diy generation expect to be able to modify and hack their devices, the way the hardware is accessible and miniaturized is important.

This means that for Digital Powertools, there are three possibilities for a business model. The first one entails using the opportunities of 3d printing to publish plans for the different tools, as well as the core software behind the tools.

Then as a way to monetize the tools, a complete set of tools, without having to DIY can be made available. This way there are a number of different ways to appeal to customers. People who are simply interested in the novel way of interacting, can just buy the completed tools, connect them to their computers, and use them out of the box.

The people who prefer to build their own software or hardware, can use the open source plans and base libraries to create either the predefined tools, or can use the plans to build their own adjustments or augmentations. The people making their own tools and especially their own plugins and software will be contributing to an ever increasing ecosystem.

The second business model is using the crowd-patronage system to promote the new tools, using the patrons to get the funding for a more classically produced small-run production line up and running, as well as getting a ready made group of people who are willing to use and develop the software. While the major driver and general purpose software will be part of the original production run, the community will be the driving factor for making specialty plugins for more and more application.

Then there is also the option of combining the two models, using the startup money to create a higher quality product line, but combining that with a 3d printed DIY line that people can modify to their hearts content. This combines the two attractive properties of the two systems. It allows for the bigger penetration and media reach of a crowdfunding campaign, with the openness and community involvement of an open source hardware approach.