Technology selection – science or art?

Distributed Systems Architecture, Foresight/Research, Frontpage, Internet Of Things (IoT)

Technology selection – science or art?

Technology selection – science or art?

In a world where technology changes by the minute, selecting a solution can be a challenge and even carry tremendous risk.

So how do you identify the right products that will serve your initiative well, gain widespread adoption, and deliver meaningful long-term impact? At Reluminati we are designing methodologies for successful technology selection in this fast paced arena.

As we delve deeply into this project over the next couple of months, we will be sharing our lessons-learned, paths to de-risk investment, understanding the moral/ethical ramifications of deployment, community engagement, innovative financing, and many other aspects of the research.

Up first: Security Camera selection. A random place to start, but it is amazing how important remote cameras have become in our work: from enabling deployment of a micro grid site in Africa, to ensuring safety and security in Haiti, security systems (cloud based or otherwise), have become de-riguer. As this piece is about technology selection, we will leave the moral implications of this rise aside for the moment (Safe to say we are tracking those debates as well).

We knew one thing when we started investigating cameras: these will be standards-based devices because, like our NEST thermostat, we view every sensor introduced into the system as just another node of what will become a smartbuilding. That narrowed it down to several thousand cameras, ranging in price from 50 to 2000 dollars.  A little daunting.

Reluminati team members have gathered requirements for the development of large, highly complex software systems like a million-line-of-code business system, and participated in technology-selection efforts for a wide variety of customers, like one building a nationwide alerting system.  Gathering the right requirements, prioritizing them, and then actually using them, is an art that benefits from both experience and industry standard frameworks and applications, and is the key to large-project success.

On the other hand, we were just picking cameras, so we did away will all that and used a simpler approach:

  1. Discuss requirements from a human, non-technology perspective. Eg, ability to see in the dark, ability to provide some information even if the network goes down.
  2. Learn about technology approaches for achieving the requirements.  Seeing in the dark is aided by certain kinds of imagers, or possessing infrared lighting. Ability to retain information without a network can be met with on-board storage.  We were even able to translate softer needs like reliability, ease of use, ease of integration, ease of support, into more concrete, quantifiable measures. Reliability could be improved through the ruggedness of the case or assessed through user-reviews; ease of integration could be indicated by support for data-standards or if the model were mentioned as compatible with other products like video server software.
  3. The final step in developing selection criteria was creation of a list of specific features,  specifications, or thresholds that traced to the approaches. For example, there are two kinds of IR lighting, LEDs and “Array” along with a few other technologies that aid in night-vision. A list of data standards was developed.

This third step came by searching for cameras that met the requirements and approaches we had even before we knew all the latest standards and grew in a highly iterative process.  For example, we were chagrinned to learn that some cameras have TF cards for image storage, not standard micro SDHC cards.  Then we learned those are the same thing.  Data standards are an even chunkier alphabet soup, and caused us to read articles on, eg, onvif vs psia.

When we narrowed down the list to our must-haves, it was easier to shop knowing that ANY camera that survived our specification-sieve would meet the user requirements, even the cheap ones.

In future articles, if you express an interest, we’ll go deeper into the technology. After all, one of our team helped write the book on video for public safety.

This is how we approach technology discovery and selection: not one more framework or spreadsheet than is necessary, yet enough rigor to ensure the right choice.




Tags: , , , , ,