Pushing a Progressive Paradigm
Encirq Opens Access

Evolutionary products are easy to launch.  You crow about this or that creative “plus-one” feature or brag about how you’ve boosted gas mileage from 25MPG to 28MPG – consumers have an easy time with that.  A one-to-one replacement for an established, existing product or capability makes for easy customer transition and low barrier to adoption. 

Revolutionary products are more difficult, however, particularly when they require a behavior change in the end customer.  Even in high-tech, the industry with the inertia and historical appreciation of the city of Las Vegas - the business that makes its living throwing out the old and bringing in the new – there is considerable reluctance for engineers to drop old ways of working and proven methodologies in favor of newer ones.  This is often true even if the new method appears to offer substantial advantages over the status-quo.

A perfect example of the challenge of bringing a new methodology to market in the engineering world is the case of Encirq.  Encirq (as we described in our earlier feature article) is in the business of providing data management capabilities for embedded applications.  The idea is simple and compelling enough – as embedded applications encounter stronger demands for sophisticated data management, particularly in areas like set-top-box program guides or user navigation of digital music and image libraries, you need more (and more efficient) data management than traditional embedded programming techniques can provide.

Up to a point, embedded developers can hunker down and “deal with it” – working to write cleaner, more efficient code with existing C/C++ methodologies, adding more memory to devices for tables and indexes, and generally doing things the same way except bigger, faster, and more efficiently.  It’s like the bicycle industry dealing with increased transportation demands by continuously working to make more efficient, more usable bikes.   Unfortunately, you eventually hit a point where you need a car.

In Encirq’s case – the “car” they envisioned is an industrial-strength database system based on established SQL programming techniques – adapted for embedded design.  In their model, you code your data-centric application just like an enterprise-level SQL database expert.  You are most certainly working at a higher level of abstraction than C or C++, and you need to have a much higher level of confidence in the underlying tools and utilities, because you know that, somewhere under the hood, C/C++ code is being generated.

The upside of this plan is that you can write less code, get faster, more efficient data management, get code that is portable to a variety of underlying hardware and OS platforms, and generate fewer bugs in the process.  The downside is that you have to trust in the software, and you have to adopt a completely new way of working – at some point and at some level, you have to switch to SQL-based application development.  Herein lies the challenge for companies like Encirq trying to establish new paradigms in embedded engineering.

Interesingly, the SQL development model has been sitting there all along, well established in sister-software areas like enterprise applications.  In the embedded world, however, “that ain’t the way we do things around here, boy.”  Encirq has responded with an innovative answer aimed directly at the problem – emotional barriers to adoption.  Encirq has just announced a free “DeviceSQL Quick Start Suite” that can be downloaded from their website.

The company has seen very high adoption rates in geographies like Japan – interestingly - where the average age of embedded software developers is younger than in the US.  By making a try-before-you-buy downloadable test drive, they allow developers to experiment quietly in the privacy of their laptop without having to bother the manager for a PO for something they fear they might not end up adopting. 

The suite includes documentation, a DeviceSQL compiler, Service libraries, the DeviceSQL prototyper, and an example MP3 player application with a tutorial that lets you get started with the technology without having to make an investment or a difficult organizational commitment.  This approach trumps even the typical “30 day eval” approach, because an engineer can safely start the process without having to engage a sales person – a process that can be intimidating for many typical software engineers in large, bureaucratic organizations.  The primary restriction with the Quick Start Suite is that you can’t use it to deploy commercial applications. Encirq generally sells their software on a royalty basis, so negotiation is required when you commit to commercial use.

Encirq has also provided discussion groups supported by their own staff, as well as a host of downloadable support collateral – all of which is intended to bring designers into the new methodology from the grass roots up, rather than by selling upper management on forcing down a major methodology change.  The company seems to have thought through the evaluation and education process quite thoroughly.

Over time, we will most likely see other companies with paradigm-bending products adopt a similarly open try-before-you-buy model.  The low-touch, low-risk, internet-based evaluation platform, especially for a methodology-changing product, seems like a strong force in winning over the mind of the engineer and creating a pull for adoption from the ground up.

Over time, we will probably also see much more widespread adoption of data-centric programming for embedded systems.  With the proliferation of highly-connected, data-driven, converged applications, a higher level of abstraction in programming makes good sense, and leveraging the work of generations of database experts in efficient data management algorithms is far more powerful than creating your own re-invented database wheel each time you start a project.  Once embedded developers have learned to trust databases in the same way we’ve come to trust compilers, the barriers will fall, and mainstream adoption will commence.

Already more companies are diving into the embedded database space, and software methodologies, tools, and IP continue to trickle down from the enterprise and desktop environments into embedded development.  Given that embedded computers actually account for the lion’s share of the computing devices out there in the world, it would be nice to see software development methodologies begin to emerge from there someday, rather than the other way around.

Kevin Morris, Embedded Technology Journal

February 6, 2007

[back to top]