The Hardware in KNX is fairly expensive. Expect costs higher than Crestron. This is industrial hardware coming into the field of HA. The hardware can cost $10 to manufacture (if that) and can be resold at $100. The margins go to support the whole ecosystem it takes to implement this in the field. The fact that the Chinese implement the hardware doesn't really change the economics. What needs to change is the cost of a proprietary channel. Crestron for example leaves 60% of its margin on the table for the integrator.
KNX hardware is known to be robust and 'just work'. Examples include the Beijing Olympics which were implemented using KNX technology.
Margins
Margin structure in KNX is similar to other hardware vendors. 10 manufacturer 40 vendor 50 installer. The big cost is in sales and distribution as well as installing and maintenance. Business models improvements can happen in both areas. Say lower cost of maintenance (remote tools).
ETS
Software suite to install, configure, manage KNX installations. Windows based software suite. Falcon is COM objects. You can plug in a laptop by way of Ethernet, USB/Serial. Once you detect a network, the system goes to S-Mode (system mode). You can set addresses, reconfigure devices, install groups. Tools is a windows based configuration application. KNX is firmware. Software written by hardware guys to offer a minimal layer of user-friendliness as you set bit-level parameters in your devices.
Annoying feature. To justify the differentiation in prices, device level features are offered and come with their own configuration sheets. This is completely non-standard.
Scene programming is complicated and device specific. A group is bits of information with a value, not many types of bits with many values and so is not a native construct.
Areas of collaboration
KNX is a standard we should support. It would offer interaction with a large body of professionals. HomeServer type tools would be interesting. Programming custom buttons at the user level could be done through the iPhone/manager tools. This is not provisioning, it requires the presence of a ORC plugged in to publish to exported actuators.
iPhone assembly. Assembling iPhone application with custom menus would be of great value for a lot of folks in the KNX world.
Standards and Business model impact notes
By standardizing a message format, and focusing on certification, the KNX organization is in effect creating an open marketplace. These dynamics exhibit self-feedback which leads to non-linear dynamics. These in turn can lead to rapid emergence of temporary monopolies.
The economic underlying principles are well known and have been seen at work in the PC industry (Wintel vs Mainframe) and the enterprise software field (J2EE vs proprietary). Without standards each vendor is left to integrate with each other vendor. For example a Panel vendor will need to integrate with a Crestron, AMX, Insteon etc. Each product integration means a hypothetical 1 unit of man time effort. For N vendors, the combinatorial means the effort and thus cost of industry integration will scale as Nsquare.
This simple combinatorial analysis explains why, in the absence of industry standards, vendor standards emerge as stable point solutions (Crestron, AMX) as the cost of variety is expensive.
By contrast in the presence of a standard, then vendors need to integrate once investing only 1 man time unit of effort. For example a panel vendor implements the KNX protocol and doesn't really care if the lighting is implemented by AMX, Siemens or Vantage. The assurance of the standard is that they will inter-work. The cost to the collective industry is N. It scales much better.
This simple analysis explains why, in the presence of standards, a new equilibrium emerges, usually initially alongside the older equilibrium, and it focuses on the operating system that enables this one standard (e.g. Microsoft), the focus is then on integration of standard pieces (e.g PC assembly, DELL).
The above is a steady state analysis of equilibrium. The dynamic picture, how we reach this steady state, is slightly different. At first, in the presence of vendor standard, a open standard will appear like one more thing to integrate with. In effect there is none of the cost advantage describe above. Au contraire, offering integration to OR for example will appear as an additional marginal effort. It should be noted that as people integrate, however, the benefits increase, as the marketplace increases, this is a positive feedback. Eventually you reach steady state where the above analysis kicks in. Again cost born by the industry to integrate goes down as the numbers increase which in return lowers cost of solution which feeds itself with more users and more choice. These self-feeding positive loop notoriously induce non-linear dynamic behaviors. This means that the trend, once started can compound fast. They are unstable dynamics that lead to rapid emergence of standards.
The meaning of this means a focus on initial marketing and sustaining PR efforts. It may also mean investments to seed the first integration in a given time-frame since the energy barrier is slightly up and then quickly down kickstarting these should be the main operational (engineering integration, marketing and PR) and financial focus.