In this talk, Erik will observe the parallels between the emergence of XP and the current landscape of IoT development. He will review the value chain for IoT development from device to cloud an comment on some of the unexpected challenges – suggesting how we can mitigate, XP style.
Sensitive viewers are warned that commercial examples of working IoT systems might be displayed.
There are huge expectations all over the world of a near-future zero-friction digital society running on cloud services and IoT devices, the Internet of Things.
The winning concepts, the blockbusters, are easy to see in a rear-view mirror, but hard to aim for, which currently frustrates entire industries (such as telco operators), investors and the stock market.
Enter the Long Tail, as experienced on Amazon.com, where hundreds of thousands of small teams make and sell niche products to millions of people who desire something a bit more bespoke or special. Amazon recently profited more (51% Q4, 2017) from the Long Tail than from anything else. Agile and open source democratized software: many small teams make software for a multiverse of applications for the many. IoT devices not so much.
The IoT challenge of building a smart, manufacturable device at the right price point is about more than just effective software. Most competing chips-and-sensors vendors offer some golden nuggets, but most try to build their own full verticals and entire eco-systems. The result is that no vendor has the traction needed to make IoT really easy. Recent attempts at just parachuting major tools and methods, from the industry of agile, has not helped.
IoT today looks like software and the web from 20 years ago, when many competing vendors tried to build locked-in kingdoms, of proprietary business models of tools and methods. A few became large, but not large enough. The big three (“The Three Amigos”) then joined forces to achieve critical business mass in a Unified Method. At its peak RUP consisted of 12 thousand documents, a plethora of potentially useful things, or golden nuggets from many places. To fully utilize RUP, each customer had to make thousands of decisions on what to use, not just once per organization, but for every project. A insurmountable effort, before you could start to make software and produce value.
Extreme Programming, XP, was a disruptive force 20 years ago. XP was just a few ideas and principles, some practices – proven by respected people, and three promising non-proprietary technologies (Refactoring, TDD, CI/CD). The XP promise was ‘Just do these things and you can deliver working software every two weeks’
IoT needs an XP-spirited approach, a similar highly constrained, cross-disciplinary team approach to rapid, iterative development of smart manufacturable inter-connected devices.
Let’s go and make great stuff – in the Long Tail of IoT – XP-style.