Practical Intro to MBED

Whilst Arduino is becoming a household word, MBED, it’s closest “rival” is not nearly as well known as yet.

Rather than use a shield approach MBED is a small DIL daughter-board with an ARM processor, a reasonable amount of RAM and flash and 0.1in pin spacing that conveniently slots into a breadboard. The IDE is cloud-based rather than downloadable. There are many libraries available, covering a range of peripheral devices and protocols. Much more detail here.

I was fortunate to attend a half-day briefing organised by the EKTN and led by Simon Ford of ARM. This allowed us to get our hands on a nice breadboarding kit for experimenters with the platform. Apart from Simon’s insights a key feature is an online cookbook with a lot of code examples which made it easy to get some quick results.

I have now done a few projects with the platform and am duly impressed with many aspects of the MBED approach.

So far, my issues have revolved around the libraries. Many of these are user-contributed and, while, mostly good enough for quick prototyping, leave much to be desired. The existence of multiple HTTPClient versions was problematic as there are several and they are not compatible. So if you find yourself using one and looking at the doc for another you are in a mess. In addition, the amount of doc varies. Some have good use case examples and others do not. At least there is a forum where you can (and I have) posted questions and observations.

The free-for-all approach that allows developers to share code and fork libraries is a two-edged sword. I reckon this could get out of control if the curation is not stepped up a notch.

I’m also aware that some people do not like the cloud-based compiler approach, preferring tool chains of their own choosing. Departing from the mainstream approach is entirely feasible for those willing to put in the effort.

Next step: investigate various prototyping boards and see whether this could beat Arduino on cost for small quantity applications. It certainly does have seem to have the edge on processing capability.

Smart Heating Control Continued

As described in my last post we are pleased to have Telepure’s “Intuition” system installed and running well. It is now running with 5 sensors around the house and I get real-time reports so I can see what’s happening at a detailed level. My ongoing task is to adjust the targets so that we are not overheating the air or water. We have seen that the system hits these targets with precision and properly compensates for the external temperature.

I have raised several points relating to the interfaces both for end users and for “energy stalwarts” like me who want confidence that the automation is delivering the results we need. Being one of the few trialists means that I get a great response to questions and suggestions. Going forward, the company is talking about a user forum and this will allow a scaleable way of dealing with the likes of me.

So far I have mainly used still-protoype web interfaces to access the system because flaky phone coverage means I generally use (cable) Internet when at home. I also like my large screen for graphs etc. However, I will also evaluate the smartphone interfaces in due course. It was good that I was able to easily construct reports that showed the Intuition’s data and other real time data that I collect on that same graph.

As this system reaches market the company faces the acid test of whether people will buy the payback proposition. However, with all the data being collected, this should not be difficult. I am building up measurements so that I will be able to talk numbers in due course.