Why Open Source with Drive IoT Innovation - Thingmonk

Editor's Notes

• #3: To start my presentation, I think it is important the we as an industry learn from our history. In many ways, we can look at the history of the Internet to see how the Internet of Things need to develop.To do look of the history of the Internet, I turn to the source of all tech knowledge, tech blogs like Tech Crunch, RWW, etc
• #4: How many people remember in 1990 when Tim Berners-Lee decided to create a start-up to commercialize some of his ideas about the Internet. The company was preparing to take on large successful companies like Compuserve and AOL.
• #5: Or in 1994 when Linux was able to land that huge round of financing to finally developer a true operating system for the Internet, called Linux.
• #6: Or when in 1998 for the Internet was when Apache finally hit $250 million revenue, on its way to $1 billion.
• #7: And finally in 2000 when IBM announced the ambitious plan to invest $1 billion in Linux. OK, this one is actually true and I will come back to why it is important fact that openness drive profit in commercial companies.
• #8: But my main point is that the previous 3 headlines did not occur. In fact the Internet was created on open standards and open source software. Hopefully it is not a too controversial statement that openness will always win over closed proprietary. This is especially true for technology that is expected to be broadly adopted. There are enough cases studies and proof points that technology based on the principles of openness make it much easier and faster for technology to be adopted. It is the transactional costs of proprietary solutions, ex bi-lateral agreements, purchase costs, competitiveness that limit adoption of technology.
• #9: There are lots of examples.Does anyone believe the Internet would be as successful as it is today if Tim Berners-Lee has created a Silicon Valley start-up company to commercialize what he called the www. There were already many successful online communities, AOL and Compuserve, that were widely used. The WWW changed everything since anyone could participate and create something without asking for permissions or negotiating an agreement with Compuserve or AOL.
• #10: In fact, the Internet is not run on open source software and open standards. Proprietary solutions still exist but it is clear who won.
• #11: Another example in the software developer area was the Web Services vs REST protocols and styles. Clearly REST has won over Web services.
• #12: Of course there are exceptions: Apple being a stellar example. However, I would submit that the Android ecosystem, which is definitely more open, is emerging as the dominate platform.
• #13: There has been a very important trend in the last 5-10 years that can’t be ignored in any technology industry and that is the importance of developers. Gone are the days of the top-down sales and procurement approach to technology distribution. Developers have become incredibly important in determining the success of new technology. If a technology is embraced by a large developer community the adoption of the technology will increase. A recent book by Stephen O’Grady documents the rise of the importance of the developer. His point is that developer are certainly the New Kingmakers for technology adoption.
• #14: A developer community will do a number of things for the technology:Increase the usefulness of the technology through add-ons, applications, and information resources.Source of innovation of how the technology can be used and applied.
• #15: Developers also excel at word of mouth promotion. The love to tell their colleagues about the next new great thing.
• #16: Key question: how do you engage a developer community?One answer: openness. Developers have no money and the time they spend investigating new technology is often their free time. Therefore, their time is scarce too. Openness lowers the barriers for developers to experiment.
• #17: Therefore, I would submit if the IoT industry wants to drive forward with innovation they need open standards and open source.
• #18: Today IoT is not open and I don’t see much of a developer community.
• #20: Lots of proprietary silos. Proprietary SDKS, proprietary protocols, proprietary solutions
• #21: We need to get to an open ecosystem of IoT technology based on common frameworks, protocols and tools. Companies needs to make their money building value add solutions on top of these common technologies. Openness does not mean no profit. A lot of profit has been made by using the common Internet technology. However, the profit has been made after a common set of Internet building blocks were in place. Ex. Apache Web Serve and Linux run the Internet.
• #22: Three trends I think will help drive an open IoT ecosystem:1. Developers are getting engaged and they will select open solutions. IoT is fun for geeks but they don’t have the money or time to navigate proprietary SDKS or protocols. They won’t tolerate it. They will migrate to open solutions.
• #23: 2. Open hardware has significantly lowered the barriers for developers to prototype and experiment. 5 years ago it was impossible to have an open source project that focused on IoT. There was no place to run the code. Developers would need to buy reasonably expensive hardware to just test their code.
• #24: 3. ioT is a huge opportunity for the big data and traditional enterprise vendors. Companies like IBM, Oracle, SAP and others see the opportunity of IoT is in the data. They will want as much data as possible coming into their systems. They won’t have the resource to do proprietary solutions for different IoT vendors. They will drive and expect a common open standard.
• #25: Eclipse M2M is initiative to create an open source community and collaboration that will enable an open ecosystem. 3 main open source project now that focus on frameworks, protocols and tools: mihini, paho and koneke.
• #26: Very important is the commercial ecosystem using the open source projects. There companies are building the open technology but also using it in their commercial products.
• #29: Paho is the reference implementation of MQTT, a very simple publish-subscribe message protocol for M2M applications. MQTT is now being standardized at OASIS and the reference implementations will be at Eclipse.
• #30: Koneki provides the tooling to make it very easy for a developer to develop, run, debug and deploy an M2M applications. It is focused on Lua and Mihini as the runtime. We will have work for Java in MTJ.
• #31: Many more projects coming….
• #32: Mihini, is a Lua-based framework for m2m device gateways. It is attempting to hide a lot of the complexity involved in building M2M applications through abstraction.