CIO Update Q&A with ThingMagic
Before joining ThingMagic (a company started by five MIT student Ashton used to teach) Ashton was co-founder and executive director of MIT's Auto-ID Center, which was created in cooperation with Gillette, Proctor & Gamble and Uniform Code Council to make RFID technology commercially viable.
Four years later, in 2003, they boasted 103 sponsors funding research at six universities around the world. It was at this point that Wal-Mart decided to start using the technology in their distribution centers.
Where is RFID technology today?
It's getting better all the time. This is a young, emerging technology. There's a lot of work to do. What's slightly unusual about RFID is that the earliest adopters are some of the biggest, most high-profile companies in the world. So RFID is emerging into the glare of a big spotlight.
Which is a good thing and bad thing. The good thing is it forces the technology to grow up very quickly and incentivizes lots of investment and innovation. So, that's where we are today. We're seeing lots and lots of innovation but we need lots and lots of innovation. It's still not possible to take an RFID tag that cost five cents and always be sure you're going to read it. But were getting much closer to that everyday.
But hasn't the Department of Defense (DOD) been using the technology for years?
Well, yes and no. RFID is a pretty general term. So what the DOD has been using is sort of battery-based tags that cost a $100. They're really small radios. The kind of RFID they're most excited about today is much, much lower cost. Sort of five-to-ten cents, if that, for an RFID tag.
How well does the technology perform at the moment?
There are a number of different factors that can impair performance and some of them come down to 'Is it a good tag?' Making RFID tags at high speed, at very low costit's still kind of a wireless network device, it's a very simple network device but that's kind of a complicated piece of electronics to be producing almost as fast you produce the daily newspaper.
So one of the areas we see a lot improvement right now is as simple as improvements in tag quality and yieldtags that work rather than tags that don't work. Then you get into 'Is the system configured in a way that allows it to work? Are the RFID readers in the right place?'
Our fifth generation reader has been developed to really ignore the interference in the radio spectrum. It turns out thats the major challenge in reading tags. Before you worry about all those water and metal (these can interfere with the reading of tags) it's what is going on in the radio spectrum that is causing problems with the system.
So tag quality, interference, system configuration, how efficient the communication link is between the RFID reader and RFID tagthese are all things that have an impact on (reading tags).
One of the things that greatly impacts whether you read the tag or not is how well the antenna works. You often want to have different antenna for a different type of product. Before, it was sort of a one-size-fits-all. A few years ago, each tag vendor made only one or two tags.
But now companies like Avery, Rafsec they have a sort of catalog of tags now and you kind of pick you tag depending on what you want to attach it to.
So there's innovation in every directionchip-level innovation, antenna innovation, system-level innovation, reader-side innovation, end users getting smarter about what they expect from the system and how they use it. So performance continues to improve, price continues to come down.
We're seeing more and more end users that, based on that, are finding their business case. I think the market's not just growing it's growing up.
Given it's shortcomings, why would anyone even want to consider deploying RFID at the moment? It sounds rather troublesome to say the least.
The promise of the technology is kind of irresistible. It solves a really huge problem: You don't know where you're stuff is.
That's why the promise of RFID is so irresistible. Once you kind of imagine what you can do if you had such a high-level of visibility into your supply chain, you get all sorts of interesting applications. So the push for RFID was really around the promise of the technology rather than the performance at any particular moment.
Who is using it now and finding value with the technology?
Nobody's obliged to jump up and say publicly what they are doing but Gillette is very public about what they are doing. They used RFID as part of the launch of their new razor system, Fusion, so they're talking about that. Wal-Mart's been pretty public about the out-of-stock reductions they are seeing. Airports and various other places. So there are a lot of adopters now that are seeing value.
Where is it being used primarily?
Anywhere you've got stuff. The challenge needs is it needs to be reasonably expensive or knowing where the stuff is needs to yield a value fairly quickly because RFID is not as cheap as we would like it to be today. So there's no business case for putting RFID on a pack of gum right now.
There's an Accenture report which claims 50% of American companies are having to adopt RFID. That may be a little high but the fact is Wal-Mart and DOD have both got mandates out saying you need to start putting RFID tags on products you sell to us. We're still at the beginning of that process.
So if you think about who supplies Wal-Mart and who supplies DOD, that's a huge amount of companies.
What is the biggest hang-up to adoption? Is it the business case or the technology?
I think there's a real challenge with managing new, emerging technology; managing change. This is not what you do everyday. CIOs spend most of their time buying things like routers, and firewalls, and databases but you would never get a CIO saying 'I don't know what a business case is for a database.'
With RFID you have these more fundamental obstacles, which are like 'What do I use it for? How do I get financial benefit?' and sort of 'Where do I begin? How do I make this thing work at all?'
One of things about RFID is it captures information you don't have right now so you don't know how badly you need it. And that's always the challenge with new technologies. So this is not a completely new problem, but it's not one the CIO faces everyday.
The lesson for the CIO is to do something and do something small and learn from it. And then do something a little bit bigger rather than trying to eat a whole elephant in a single bite.