I have long been a proponent of the concept of do things differently. Why? Well first off because there are so many different ways to look at any problem that only using one problem solving model seems risky to me. But the other side is that I often see many potential solutions. That makes me wonder if in fact there are more potential solutions and I continue driving towards those. It is why in the end that I am not a project manager and instead stayed an Architect.
Really good project managers will release their architect and say go figure out these tough questions before they delay our project. That partnership between a great PM and an Architect can be spectacular. Right after both parties agree that it can be done, finding the best way to build and deliver the solution (and manage that delivery) is incredible.
A good PM/Architect team should combine the strategic view of managing the money and time of a project and the strategic view of how the solution should in the end be deployed. Mix into this combination the day to day operations of a team and someone looking ahead at problems that haven’t occurred but may yet occur.
Sometimes organizations pull architects too early from the project. The deployment team begins to gear up and the architect moves on to other projects. I have long been an advocate of having numerous check points where the architect re-engages with the project team to make sure everything is running smoothly. Additionally this allows for reality checks in the architecture, i.e. we can’t do it as designed the product doesn’t work that way or there are firewalls and other things in the deployment path that prohibit the solution as designed. The design and then be modified to reflect the new solution and there won’t be an automatic gap the next time the architect engages.
Why dredge up this golden oldie about architecture now? I am watching the iOT world explode around us. First off as I published on this blog a couple of months ago that I was more concerned with the data of the internet of things. (My DiOT posts). The data produced by the internet of things presents two distinct problems. The first is today the data of the sensors is controlled by the company that makes the software that controls the sensor.
I am an acknowledged weather geek. I have two weather stations at my home. My NetATMO system that allows me to connect to the weather sensor from my iPad anywhere I am. The data is published to the web in an easy to consume format. They have linked all the people with NetATMO systems so you can even see what the weather is at the house next door. The other system I have is a more traditional weather station from Davis. It has a remote sensor pole and a base unit. Using the Davis software you connect to the station and download the data.
You cannot however create analytics comparing the temperature and wind at my house, my neighbor’s house and between the two different weather stations I have. That in the end is the growing reality of the data of the internet of things (or DiOT as I call it). It isn’t just weather stations and it isn’t just two or four sensors in the end this is going to be an interesting problem.
It is the preverbal hallway with 100 doors. The door you open may or may not be the right one. You get 3 guesses. So your probability of success is low. There are many standards and many architects building the sensor solutions but nobody owns the overall impact.
Someday this may become a problem…How do I analyze the data I collected from five different sensors and create a merged view of that information…