Are we facing an engineering crisis - in other words, do we have sufficient human resources in order to build and maintain the Internet of Things? At the recent IoT Design Conference in London, Gary Atkinson from ARM suggested there could be more than a million new device developers by 2020. From the perspective of a chip manufacturer, this means that their devices are probably no longer used by just a small cohort of highly specialised electrical engineers, but that many more people with a diverse set of career trajectories are going to start using ARM chips in their connected products. It's therefore not surprising that an unique selling point might be 'security on chip', thus that the user is shielded from this level of complexity.
How much should an IoT engineer know, and/or is there even such thing as an IoT engineer? I believe seasoned engineers may disagree, and the fact that especially in industrial settings people have successfully connected machinery to the Internet for more than two decades now speaks for itself. However, I agree that IoT brings some technical challenges that are not necessarily part of the classically divided training for electrical engineers, software developers and product designers. IoT is a highly interdisciplinary area and many problems require a wider perspective. Unfortunately, in my opinion and maybe unfounded, the increasing specialisation already at the University level may further decrease the ability of graduates to easily interface with other academic disciplines.
Two recent articles stated that IoT itself is so vast and complex that pinning down a single skill is certainly not easy (zdnet) and Computer Business Reviews listed five skills necessary to succeed in IoT development:
- Computer programming
- User interface design
- Business intelligence AKA analytics
- Hardware engineering
- Artificial intelligence knowledge
I don't disagree with that list and would even extent it with knowledge of IoT architectures, cloud platforms, distributed computing, security, data protection, etc, etc, etc. In fact many higher education institutions are already gearing up to offer or are already offering courses in IoT. The visionary IoT School even goes a step further and offers experimental kits for key stage 2-4 children.
Just as putting a chip into an everyday device doesn't make it a good IoT product, simply teaching an electrical engineer a bit of analytics or educating the user interface designer about power saving mechanisms doesn't make them promising IoT developers. What is required is an integrated approach that makes them appreciate every level of this complex ecosystem.
A few months ago I suggested how a final exam for an aspiring engineer might look like:
The tweet was well received and a few people even said they'd like to use it for interview questions. Only then it occurred to me while 'practitioners in the field' have accumulated the relevant expertise over the years and you can find answers to most questions in various text books, I don't know about a single book that would help people learn the necessary knowledge by reading it cover to cover.
As many readers may know I'm formally educated as biologist. My specialisation is in a particular branch of molecular biology and for my final examinations I was expected to know the contents of various text books with titles along the lines of The Molecular Biology of... or The Biochemistry of.... However, we had a guide book that introduced the basic principles of chemistry and biology in about 1,400 pages, and although it only scratched the surface of most subjects, it always served us as orientation and to obtain the necessary vocabulary to be able to appreciate the context for our more specialised reading.
I'm missing that book for the technical aspects of IoT. Sure, there are specialised books that help to design wireless devices, those that highlight monetisation strategies for IoT, and a flurry of titles "IoT with..." that provide hands-on solutions for a particular microcontroller platform, programming language or web frameworks. I truly appreciate those books as I've learned a lot from them over the years, but they are far from a focussed text book that covers the technology of IoT conceptually. Also, from somebody who graduated from a higher education institution, I'd expect to have more than a vague idea what the theory behind that technology is. A book on IoT circuit design that introduces electricity as god-given force doesn't go far enough. IoT (as well as any other technology) is the clever application of the laws of physics!
I've started to write that book. It is daunting as I'm obviously not an expert in everything that is relevant to IoT. My draft table of contents currently has 109 headings, many of which make me shudder: Talking about wireless just requires an introduction of Maxwell's equation, and if you're thinking about IoT platforms you should have heard about the CAP theorem, to just name a few. But I believe that "B.Eng. IoT" students deserve a good book and that graduates need a solid foundation that leaves them with deep knowledge, and not just as implementation monkeys.
Please get in touch if you're interested to help (badryan@gm...). I need contacts to publishers, I need proof-readers, I need people with domain knowledge of all aspects of IoT that I can ask. Quite probably, I may also need a few co-authors to make this book happen in a reasonable timeframe, i.e. during the next year. Also, if you've written a technical book before, I may have a few questions...