Monthly Archives: September 2009

The Elephant in the room – why custom software solutions are almost always bad news for RFID

Despite the numerous advances we have made in the areas of hardware capabilities (pertaining to performance, capabilities and costs – see “The changing landscape of RFID – Then and Now”), the big question mark that still hangs over us is around the availability of scalable software solutions that address end user needs – in other words:

What are the factors still hindering the rapid adoption of RFID – especially since the value to end users are well documented for a long time now?

The biggest risk for the end users today as I see it is the fact that to date most solutions are all custom developed from the device layer up – what does this translate into for the end user? ….Consider the following risks:

  • Custom solutions often requires wide range of skills to deploy a working solution (RF, device specific, business process consulting, connection to enterprise applications, etc.)
  • Custom solutions do not ‘evolve’ easily as technology advances– and hardware technology as we’ve seen continues to evolve by the day!
  • Custom solutions result in high risk and high on-going support

These solutions are normally point solutions that are difficult to implement, deploy, resulting in numerous issues in ability to distinguish Hardware, Software, or System design errors when they are deployed in production (i.e. troubleshoot, maintain and scale).

At the end of the day, you are left hanging at the graveyard of unsuccessful ‘process re-engineering’ projects that cost too much, are hard to implement, don’t really work, and look something like this…



Source (elephant and blind men picture):

Perhaps most scary is the aspect that most (if not all of these solutions) are often found trying to replace an existing Line of Business (LoB) layer vs. extending and complementing the existing electronic system to be ‘real-world’ aware i.e. change in process at the same time as change in technology is almost always a bad idea

In other words, enterprise applications have been shown to maintain a precise view of the physical world they control. However, it’s also been shown that they do not have visibility into the real world events [a great example is the lack of visibility in warehouses in terms of what the WMS system has as system of record, and what’s actually on the floor]. It’s this loosely coupled nature of interaction between Electronic Systems and Physical World that results in a big gap between Actual vs. Expected, and if we optimize the solution to solve these requirements with a packaged real-time visibility system, we’ve avoided the above pitfalls altogether.

In summary, the real barrier (to deploying real time visibility systems today) has been the lack of packaged software solutions that connects all the parts and get them to work in concert with the enterprise applications. The only option today unfortunately (as illustrated above) all along has been to develop a custom solution from scratch which, as all custom development efforts, can be a time consuming and risky proposition for a typical end-user customer to take on… What is needed today is a solution that can be easily configured, and deployed like any other enterprise business application to gather and utilize real-time data with minimal or no disruption to existing enterprise applications.

Stay tuned for my next post this week where I will illustrate a proven approach to address the above issues.

Until then, Cheers!


Next post: The Elephant Movers – S3Edge’s approach to systematically break down these barriers and deploy a scalable, enterprise-wide RTVS

Digg This

The changing landscape of RFID – Then and Now (with recent enabling developments)

Then: Let’s be honest – Hardware was frankly, a big drag – Readers were big, bulky, consumed a lot of power and extremely hard to deal with, and tags just did not meet production requirements (more of an art than science to get to the magical 100% read rate and the cost factor was almost always an issue in terms of justifying investment)

Now: We are starting to see the early ‘iPhones’ of Hardware making their entry into the market – Specifically we are seeing devices / tags with the following characteristics

  • Better and smaller form factors (~ 75% smaller)
  • Way less power consumption, thanks to Power over Ethernet or PoE (~50-60% lesser)
  • And a whole lot cheaper… yep, sometimes upto 40% cheaper (!)

Here’s some anecdotal data to substantiate this:

Impinj announced Speedway Revolution fixed readers; $1,385 list price for a 2-port and $1,585 list price for a 4-port {Motorola FX7400 is similar}

–Convergence Systems introduced its CS203 integrated reader/antenna; has PoE, and rated for use in extreme environments, has passed Mil Std 810 testing, and has a list price of $700.

–Lexmark introduced its RFID Gen2 UHF Option for its T654 monochrome laser printer; list price of $2,499

–UPM Raflatac introduced its Dogbone Freeze tag designed for temperatures reaching -40C

–Impinj Monza3 available in packaged format for PCB integration

Then: Wireless technology / devices were unreliable and expensive to configure in concert with RFID (i.e. product capabilities, deployment knowhow, and cost of ‘active’ tags were all issues)

Now: Wireless devices are now ubiquitous and low cost; Reliable wireless networks (specifically UWB, GPRS) that allow for RFID data to be captured within the  4 walls of your enterprise in a seamless manner (providing continuous visibility of high value assets in a specified area) are now more widely available – Some really neat technologies to look at here are Time Domain and Sandlinks which utilize UWB in neat ways to provide continuous visibility – In addition, costs associated with tags to roll out these systems now scale in volume which is critical for adoption (vs. trying to impose a goofy per tag licensing model to remunerate location software costs!)

Moreover we are starting to see mass proliferation of low-cost intelligent mobile devices that are wirelessly connected and work reliably in form factors applicable to the target work environments (ruggedized handhelds, vehicle mount computers, fixed readers etc.) – perhaps most importantly software now can take take advantage of computing resources to implement significant workflow processing on device thanks to the advent of mobile computing platforms in the form of BizTalk RFID Mobile.


What does all this mean in layman terms?:

For the fist time it’s possible to cost effectively address operational issues – that result in errors and decreased utilization of assets / people – by deploying systems to automatically track and report on material movement events in real time.


So if the above is indeed true, what is preventing end-users from taking full advantage of this technology and utilizing it to add value to our environments?

Well, there still remains (unfortunately) a number of current barriers to adoption – especially from a systems standpoint in making all the moving pieces (hardware and software) work together in concert for you – that I will talk to in our next post.

Look for my next post on this topic first thing Monday.. until then, Cheers!


Digg This