The TCM chain goes electronic for better all-round service, as ALFRED SIEW finds out
Article on Straitstimes Digital Life 4 Dec 2007

It is not an easy task to profit from such a project, from the point of view of a software application developer. That is the reason why big players in SI System Integration are not keen with such customers. The reason being the Long Tail after the project commissioning. It is the choice of Eu Yan Sang and half the reason could be due to the cost. I'm not sure of the track record of Philip Tang and Sons in SI projects, but I have knew of several local startups serious in SI projects could not last very long doing the same specialisation. Some of these startups are fresh graduates from NUS who have not entered the working industry yet. The feedback from these cohort are usually that they are not expecting a Long Tail in SI Projects. If they had known about this Long Tail, they would not have chosen SI as the trade. Academics and Products Pre-Sales painted a very rosy picture about SI projects. In conclusion, Long Tail is a must in SI projects and software developers must be able to swallow it completely unless the customer undersands about such issues. Usually companies who had tried and tested would appreciate it. But that does not necessary meant that companies are willing to pay for it. Some may think that it is to make it pretty, but the fact is just to prevent it from turning urgly.

This is a great innovation! You do not need complex RFID technology for monitoring simple beverage in a small store like a minibar in a hotel room. Just simple cmos sensors at strategic location should do the trick. Simple is solution!

As the name implies, the shelves in National Library will soon be very intelligent. It should be able to "sense" if books are on the shelf or have left it, and how frequent it has done it! This may not be something that impossible in this current era, but the interesting part is that it is a non intrusive solution.

When RFID technology arrived, many see it as an advancement over barcoding. We need to scan items individually, making sure the face of the barcode and be visually seen by the reader. Then we now have RFID which also need certain range and line of sight before the reader can "sense" it. The major improvement is that it achieves an "Invisible Force" when doing the reading and detection, which is far more non-intrusive to the humans using it.

After talking to an RFID expert, I realised that there are many challenges to achieve such a simple requirement using the  RFID. The radio frequency emitted by the tags on the books are strong enough to penerate the paper and shelf thickness. The application of such a technology may not be possible when the items to tag are made of different materials, such as liquid which may be anti-radio frequency. As we all know, the Live Radio Station Air Broadcasting technology we have been using for many years, have problems of reception if blocked by tall buildings.

Other than working on the hardware technology of RFID, the subsequent challenges would be to get it integrated with its existing software applications and databases. Apart from the ability to charge up the Tags, which emits radio frequency signals to the reader, an intelligent software is also required to be put in place to determine which tags is who and when is the signal sent, and what is the timeout for a retry.

In conclusion, there is quite a lot of challenges, before this technology can take place successfully. This is good news that Smart Shelves RFID solution have taken a big step towards Trial. It is just a milestone to be recorded down in history, not something for us to celebrate yet. I believe this technological advancement needs both Hardware and Software solution tightly integrated to be successful. Its not just about hardware, but quite a big amount of work in software too.