Thingscon, a conference that fosters the creation of a human-centric & responsible Internet of Things, was held in Amsterdam from 30 November to 1 December 2017. The event was attended by a diverse group of IoT ecosystem operators, from companies to researchers and from designers to developers. Also in attendance were Program Director Hanna Niemi-Hugaerts, who took part in the panel ‘How to develop a Citizen centered IoT’ (panel report), as well as IoT developers Aapo Rista and Henri Kotkanen from Forum Virium’s IoT team.
The conference has been successfully organised in several different parts of the world in recent years. According to the conference’s opening speech, the content of the event has reflected the development of IoT all the way from silly little gadgets to the present expectation of all devices being networked. This year the conference highlighted major challenges related to the full-scale utilisation of IoT, such as the need for artificial intelligence, the growing social impact of algorithms and the requirements imposed on devices by the GDPR.
The Forum Virium IoT team have picked out the five most interesting topics from the conference, which we believe will be of interest to anyone working with IoT.
1. Rules for IoT development: IoT Mark
IoT Mark is a certification mark that is striving for higher quality IoT services and devices, consisting of thirty principles focusing on areas such as privacy, data security, product lifecycle and interoperability, among others. The principles remind manufacturers to make sure that devices function even without an Internet connection, that device ownership can be transferred and that devices can be repaired. Recommended reading for anyone wanting to introduce more ethical services and devices to the market, whether you’re approaching the subject from the perspective of data processing, service quality or manufacturing.
2. How to get started with machine learning: Useless Butler
The purpose of this workshop was to showcase how easy it is to start using machine learning in your own projects. The goal of the two-hour session was to build a simple connected product (out of an Arduino MKR1000, 2 light sensors and an RGB-LED) and teach it to react to light sensor readings – not the easiest of tasks. The majority of the time was taken up by the introduction and connecting components with jumper wire, but in the end almost every participating group was able to teach their ‘useless butler’ to react to light sensor changes by changing the colour of the LED.
The teaching was conducted using open source Wekinator software, which provides a GUI for common machine learning algorithms. The software can be taught discrete classification based on input values (‘turn on 1, 2 or 3 lamps based on the amount of light outside’); regression, meaning continuous change dependent on an input value (‘continuously adjust lamp brightness depending on the amount of light outside’); or to recognise patterns over time (‘turn on a lamp when you recognise two claps, increase brightness when you recognise three claps’) with just a few simple clicks (once initial configuration has been sorted out, that is). The FVH team successfully taught their own butler to light up green in bright light and red in darkness. Quite simple, but it was quite a surprise to learn just how easy it was to teach!
We also recommend taking a look at the Google Teachable machine website, which illustrates the concepts of machine learning nicely and allows you to teach your own web browser.
The workshop instructions, presentation slides and circuit diagrams.
3. Enforcement of GDPR begins: is the Internet of Things ready?
The EU’s General Data Protection Regulation (GDPR), which was approved in 2016, introduces changes to data collection practices and services that utilise data. At Thingscon, participants were reminded to keep in mind that audio recordings as well as data related to the use of a service or device are also considered personal data. The event also showcased tools with which companies can independently examine the role of data and their data handling processes and be ready once the GDPR enters into force on 25 May 2018.
If you want to check or increase your familiarity with the GDPR as regards the definition of privacy and personal data, the Privacy literacy Survey offers a survey-format tool for doing just that. Meanwhile the Information Asset Inventory is a good way of getting started on defining data handling processes for anyone who collects data in some form. Step-by-step instructions for the inventory are available in the session presentation (slides 41–56).
4. Making the algorithm that allocates car charging times transparent: Smart Charging Station
The current situation is that the electrical grid of a given street cannot charge the cars of all of the street’s residents and visitors simultaneously, meaning that charging times or rights must be allocated between users, preferably in a fair manner. In response to this, energy company Alliander and innovation incubator ElaadNL developed the Transparant Charging Station concept, the aim of which is to make the logic by which charging time is allocated as transparent as possible. The concept was emphasised as being an invitation to discuss value-based smart city development, as transparency is not the only possible democratic value on which service development could be based. The topic was also covered in the workshop session Democracy by Design, which explored what software development based on values such as freedom or equality, for example, would be like.
5. How can residents be involved?
Human-centrism was also present in the showcased projects. Mara Balestrini from the company Ideas for Change summarised her experiences with the Bristol Approach, in which challenges are identified and scoped and solutions to them are created in collaboration with residents.
In Bristol, the approach helped identify the problem of damp housing. This led to the development of a ‘frog sensor’ for collecting damp level measurements from residents, with the sensor’s cute appearance making measuring easy-to-approach for everyone, regardless of technical know-how. In Barcelona, the approach was used to solve the noise problems of Place del Sol. In the Making Sense project, residents were able to make the problem visible through measurements and convinced the city to make adjustments to the urban space that mitigated disturbances resulting from night-time partying in particular.
You can find recordings of the sessions here