Thiele’s Law and How it Applies to Edge
First thing you’re thinking is “there must be another Thiele, cause there’s no way Mark has the ego big enough to think he can have a law named after him”. Funny story, but I claimed this “law” a while back on LinkedIn and since no one argued, I’m keeping it! The law states
“In technology, new services enabled, quickly become services required”
OK, maybe I do have a big ego, shoot me!
How does Mark’s ego have anything to do with edge computing?
As each day goes by here at Edgevana and via the mediums of Twitter, virtual events, email and LinkedIn I gain just a little more knowledge about what might end up as an Edge solution and what it will take to support it. I’m seeing companies of all shapes, sizes and industries poke and prod at the edge. In many cases the first applications are delivered via some on-premises hardware and software stack in a factory or in retail outlets. These applications vary from safety on the shop/factory floor to customer experience in a store. However, they often have one thing in common, they are enabled in a “this is nice to have” type of way. In other words, use the app if you like it or don’t, business will go on. However, what quickly happens is that the apps move from nice to have, to required. Hence, Mark’s big ego and the phrase.
“In technology, new services enabled, quickly become services required”
What does this mean for “Enhancements” to a facility or operation somewhere?
Favors can bite you when you're in IT. However, when doing a favor, you have to be prepared for owning long term support and resiliency of that favor. Think about a large retailer where an application has been enabled to speed customers through the store without the need to checkout. Now, think about the staffing changes that are likely to be made over time to adjust to the reduced need for checkout clerks. What happens when the automated checkout doesn’t work, you have a ton of unhappy customers attempting to make their way through a smaller number of checkout clerks. The aforementioned example is fairly obvious and would likely not be enabled as a “favor” to the store. However, there could be a dozen little applications that when combined, eventually create an always-on requirement.
Considering the numbers at the Edge
Other considerations when deploying new services to the edge include but aren’t limited to;
- How many locations?
- How many interactions at each location?
- Where are the locations?
- What type of support do you have available?
- How easy is it to push upgrades, patches and security updates?
- Are there potential add on effects that local users might enable? (I.e. Combining two applications in a new way)
- How much data will the applications create and what’s the value (I.e. How will you need to treat that data?)
- Are the new capabilities a privacy or security risk, could they become one?
- What processes have you put in place to capture new innovation on your application that should be shared with other sites/all sites
- Should you make one of the applications a training application to reduce helpdesk tickets?
- What type of resiliency does the application require and what are the best measures to weigh in making the correct determination?
- Have you focused on locking the applications down or creating an environment where local experimentation might occur?
- Etc., etc..
Long story short….
Thiele’s law was created because I care. OK fine, it’s my ego. However, it doesn’t change the fact that a small thing delivered remotely might go from novelty to minimum success requirement for a site, every site, a group of employees or your customers. Don’t be too generous handing out puppies (applications) unless you have a plan for helping to pick up the poop.
The Edgevana Blog: https://www.edgevana.com/blog/
Author: "Lean Entrepreneurship: Innovation in the Modern Enterprise" & "The Innovative CIO", Proven Transformative Leader, Private Cloud Pioneer, Team Builder, Strategist
4 年Great points, Mark. I've seen this type of thing many times. A cool prototype or PoC running on a repurposed, aged-out machine becomes "mission critical" for someone. The team creating the system may not know someone's using it in a production context. Then it gets "interesting".
CIO Strategic Advisor at AVOA
4 年I love the idea. How do you instill value into the equation? I'd worry that it may come across as a 'build it and they will come' approach. Value would change that.
Tech geek, father, photographer, occasional author, notorious riverboat gambler. CISO of the best company in digital healthcare. I build and secure stuff that scales.
4 年Remind me sometime to share with you Dougherty's Law.
Unlocking Real-Time Insights for Utilities in DR, EE and DER | Driving AI & Automation in Manufacturing & Mining
4 年I need to start writing some laws too ;-)