The IoT and the Technical Information Challenge Stuart Mendelsohn 1
About me From UK (Manchester) 4 years learning how to dig up dead people (archaeologist) Electronics was a hobby and then studied electronics/computing Started as a Aerospace technical writer (Airbus A320), Phillips, ASML, Sagantec, IBM, Ericsson, Intel, Nokia 1994 6 SGML manual Ericsson AXE (Arbortext Publisher) Arbortext/DITA wiki/knowledge base for IBM at IKEA DITA inspired wiki for machine learning/big data IoT project for monitoring servers. Currently at Tennotech a bridge between facilities management and IP video cameras (3000s +!), automatically displaying cameras for each alarm/incident. 2
Introduction Note: This presentation is about one CMS segment, The Internet of Things (IoT) that is not yet using DITA based technology I am going to suggest a definition of the IoT (Internet of Things), then briefly survey the technology. I will outline the challenge to create technical information for many of these products and technologies. Then I will explain this puzzle AWESOME WICKED WIKID but first... 3
What is the IoT exactly? By Wilgengebroed on Flickr 4
Connecting Devices to a Network or Networks Connecting objects/devices/animals to a network or networks, wired or wireless or a combination of these. 5
That is a lot of devices! Trillions of devices are predicted, so it will effect you! It is not a standard and can include anything from vending machines, drones, toys and medical devices, to mobile robots and smart city sensors. 6
Connect this! 7
Smart Cities Smart Cities are an example of an IoT market segment, the ITU has a smart city initiative. Standards that leverage IoT technologies to address urban development challenges. How long will this standards process take? What will the effect be? 8
Complex set of technologies, many very immature There are several wireless standards which can be used to connect devices to a network, most are still developing (Things network and Sigfox still being deployed, for example) That means delays as products play catch up with new networking standards. Proprietary standards too, checkout Wittra. 9
Risk High risk means you will get low cost, short projects with minimum documentation and the maximum use of open source software/hardware. GitHub open source projects The place where over 12 million read and write their technical information, via a browser interface. 10
Fragmented marketplace No clear market leaders, many brands will not last. Many products targeted at the maker community, not the end consumer. Software and hardware suppliers are hoping that makers/inventors will use their products to create mass market products. The varieties of technology in use will restrict the amount of experts available for some technologies 11
Regional IoT Markets We may see regional IoT market segments in China and India and for example, due to the size of their domestic markets. Countries may also focus on IoT technologies that complement domestic industries (IoT and telecomms in Sweden). 12
Competing wireless technologies Just two examples LoRa Technology and Sigfox LoRaWAN is a Low Power Wide Area Network (LPWAN) specification intended for wireless battery operated things in a regional, national or global network Sigfox Global cellular connectivity for the Internet of Things, fully dedicated to lowthroughput communications. Planning to be the global network provider (monopoly) for this technology. 13
Variety of operating systems Examples of Operating systems that run IoT applications are: Android Brillo ios Contiki QNX Linux Some may already be on your internet device! 14
Variety of application development environments Arduino, Eclipse, Mbed, NodeRed these are for hardware developers and are documented online and frequently updated, very technical audience, makers and hardware designers. No standard, most of these environments have multi vendor support. There are many more niche products. 15
So, Few real standards yet What is the standard way we can connect our phones to a network? Let's do a quick survey! How may ways can your phone connect to a network? 16
Remember you have a camera and a screen! and what else?. 17
Sensors, huge variety and cheap, but interoperability is a challenge From CO 2 to pulse rate, from magnetometers to flow meters there are thousands of different devices you can use to measure and interact with the real world. But there is no standard way to allow different platforms to cooperate even if their sensors can measure the same thing. 18
Sensor Standards are Needed Two robots are connected, both have compasses but from different manufacturers. If one compass fails can the other robot communicate which way is north? N? N! 19
Emerging security and identity management standards/techniques Since the ways to connect devices are still being developed, the security threats and solutions are developing too! Few if any security standards are in place yet. Security is and will be a major concern in the IoT, due to the complex frequently changing technology landscape. 20
WiFi is not secure! WiFi is commonly used in connecting devices, but it is not seen as a secure standard. The FOI, the Swedish Defense Research institute does not use WiFi! New Low power longer range (1km) WiFi standard HaLow 21
Legal landscape of IoT data different in EU compared to the US, for example. Beware of the cloud you might be storing EU citizens data outside the EU! So... Swedish optional patient's personalized health data wallet will be stored in an EU based cloud solution (using Cap Gemini's cloud servers in France). 22
The European Union initiative Regulation setting out a general EU framework for data protection Directive on protecting personal data processed for the purposes of prevention, detection, investigation or prosecution of criminal offences and related judicial activities. 23
Relationship to other buzzwords cloud, AI, Big Data. Cloud Your device might have limited capacity to process or store data so you might use the cloud. But it might not work in real time, or be secure! What happens if your driverless car enters a tunnel? Big Data With trillions of devices producing data you will need Big Data platforms and software to process the vast amounts of data involved. Techniques used by Google, Facebook and Amazon, such as Hadoop. 24
Machine Learning and AI techniques are already being used to classify and analyze data at scale, this might be impossible or too time consuming to be done by humans! AI techniques are used to identify anomalies in data from similar devices. 25
Some IoT Limitations The IoT is not a well defined mature set of technologies. This makes for a very dynamic information collection and production environment, with often minimal resources. Mistakes have and will be made, projects will fail. 26
Fast developing hardware platforms, hardware documentation challenge. This is how some hardware developers describe their documentation Being brutally honest here, we are not teachers, technical writers or particularly adept at writing documentation that encompasses the experience level of every kind of person that could buy from us. Ciseco 27
Challenges of Open Source development and documentation. wikiunab.wikispaces.com 28
Teams and Wikis Teams are often international and the only common tools are a browser. You may never meet the engineers you work with, you may see them online in a video conference, but not always. Open source software platforms usually use wikis and Github. This is common at universities too, so even more developers will use wikis in the future. 29
Technical Information Challenge Many of the products/software you use will only be documented online in a wiki, for example Github. Often important information will be in a comment or a reply to third party. Project lifecycles are short and resources are scarce. The only way to enter information is often via a browser. 30
Conclusion IoT products and the technology they use are still emerging. Collaborative developments are common in the often open source, crowd funded IoT world. Crowd funded projects on Kickstarter, for example often use wikis. The Things Network, a rapidly spreading IoT radio network project, went global in six months! Agile environments often have international collaboration, a browser is often the only common tool. Many projects do not initially translate to other languages. Wikis are the most popular collaboration tool in IoT software/hardware development documenting products. 31
My Experience Created DITA wiki/knowledge base for technical support staff, because they found it hard to understand DITA and had little knowledge of technical information or CMS systems. After the previous assignment I created a product wiki based on the three DITA topic types. Each wiki topic was assigned to one of the three DITA topic types. Procedural (how to) Concept and Reference. A simple approach that developers could understand (using examples for each topic type) and use without training or special tools. 32
Answer to the puzzle The puzzle was to explain: AWESOME WICKED WIKID Awesome in US English has an equivalent in British English...wicked! Use a subset of DITA in wikis, so Wiki/DITA = WikiD Create an open source wiki with DITA elements, even Github? 33
But! This is not a universal approach for documentation, but only for the agile (it is growing fast), often under resourced world of IoT and similar product development environments. If you have the resources and more stable projects other documentation strategies and products may work best for you! 34
Thank You! Questions? 35
Further Reading International Standards ISO ISO IoT report ITU ITU Smart Cities Wi Fi Alliance US IEEE EU ETSI Legal Issues US EU 36
Contact details stuart.mendelsohn@gmail.com +46738555302 Skype id: stuart.mendelsohn Stockholm, Sweden 37