• rely on the computer connectivity to convey data( PC, Mac,…that is ).
• rely on the router ( by the way, check out MyOpenRouter)
  

In either case, a piece of software and a connectivity add-on will have to be installed on the gateway.

Please refer below to the diagrams below explaining the two designs.

I think that the web of things will very likely bypass the personal computer and will leverage routers for connectivity and mobile handsets for user interface.
This design should dominate the market for two reasons:

• energy consumption: routers are much less energy-consuming than a PC. If one appliance needs to remain on
• browser is the new desktop: energy management softwares are always implemented as software-as-a-service. Therefore the computing power of the personal computer will not be utilized.
• mobile application: It’s never been so easy to provide application to monitor the home. I think that every environment monitoring products will come along with a mobile application to download.
• reliability: a router is always on, always connected and more importantly dedicated to one task
• router usb connectivity: the router provides a way to get enhanced in functionality thanks to usb dongle. The usb dongle can be the piece of software and the connectivity add-on to communicate with things around us.

In conclusion, the web of things will definitely foster the home router breakthrough as the centerpiece in smart home.
Most importantly, check out MyOpenRouter to understand the future of routers and how they can be tailored for the smart home.

“Web of things”,  “Internet of things”, “pervasive computing” are more trendy to describe this idea to hook up inanimate objects together, make them communicate, plug them to the web.

From now on, I will use the terminology “Pervasive Web” to describe this intent and this wonderful opportunity to push further the Internet frontier.

Out there, Web 3.0 is also a terminology that gets more and more momentum. It brings more confusion than clarity since web 3.0 can apply to various technologies – semantic web being one of them…and I don’t see any bridge between semantic and pervasive yet.

Another selling point to use “Pervasive Web” is the different working group trying to expose sensors as RESTful API ( check out the great Web of Things blog )…..this is exactly I’ve been implementing for month…stay tuned!

I’ve stumbled across many wireless sensors product. I’ve quickly shunt aside some of them:
- Arch Rock: no starter kit available. Their vision to bring the IP world to the sensors is pretty cool. It should open new doors for innovation and will bridge the sensed reality and the web…but for quick prototyping, it is not suitable.
- Dust Networks: too expensive for the starter kit ( don’t want to disclose but it is more than my budget ). It is not tailored for quick prototyping due to its complexity. However, it may be a great solution for robust, reliable systems.
- Sentilla: WOW, Java support, 199$ for the starter kit including two motes, one USB gateway. A mote is out-of-box equiped with a acceleration, temperature, volt, til sensors. However, it looks like their new strategy is to focus on an energy management solution for data centers. Should they drop their perk kit support, I’d rather be very careful with my chersihed motes.
- Arduino family: yes, Arduino can come with an XBee module . Arduino is definitely promised to a great future!

My decision was to go for the Sentilla perk kit. I just love the Java support which should smooth up the development phase for prototyping….I’ll be trying out the Arduino XBee module after the prototyping so as to bring down costs and schlep my demo to to the open source completely.