The large turnout showed that this is a subject that inspires plenty of interest and confusion. Maurice van den Akker, the team leader of Wireless Innovation at SURFnet, and Paul Dekkers, a LoRaWAN and eduroam expert at SURFnet, explained in their introduction that SURFnet has long considered the question of what LoRa as a network technology and IoT as a concept can contribute to the core activities and processes of education and research institutions.
Maurice van den Akker (SURFnet) tried to offer more clarity about where IoT originated and how it can be properly defined, but also about its potential for disruptive impact on both the corporate world and society. Since investments running into billions are expected to be made in the IoT for sectors such as agribusiness (smart farming, monitoring your cattle shed remotely), retail, transport and industry in the next year alone, it is clear that this is no longer just an emerging technology and the hype phase is well and truly over.The issue of defining IoT was addressed by most speakers, yielding a variety of answers. Does it refer to the network? Or to the “things” themselves such as the sensors used by the network? Does it primarily refer to the massive amount of data that IoT applications – often also grouped under labels such as smart office, smart city or smart car solutions – provide? Or to the big data component: how you extract information from raw data and combine it with other data to achieve new insights?
Dutch presentation: The Internet – Also extremely suitable for things (Maurice van den Akker, SURFnet)
IoT as a network
The answer, as so often the case: “all of the above”. But you have to start somewhere, so why not start by choosing the most appropriate network? Paul Dekkers (SURFnet) listed the benefits of using an LPWAN standard such as LoRa (Long Range Low Power) or LoRaWAN to support IoT applications.
LoRa services basic requirements better than anyone else, such as:
Wide range (in a test set-up on the roof of the Utrecht University of Applied Sciences, the antenna appeared to be able to receive signals from Amsterdam).
- Relatively cheap hardware (just a few thousand euros for an antenna with an effective range of several kilometres and a high penetrating power).
- The capacity to service several thousand devices and sensors per antenna.
- Low energy consumption by devices that only require a small capacity (small data volumes sent and maybe received a few times an hour, day or week).
Dutch presentation: Internet of Things (Paul Dekkers, SURFnet)
IoT as application
SURFnet had previously looked at the network, but an increasing number of questions from its constituency were about its potential applications. A wide range of applications were covered in the seminar. For example, using GPS trackers in combination with mobile apps in healthcare, for people with dementia. One example was presented by the applied geriatrician Margo Emmen: the Wuzzi Alert.The concept: helping to relieve the burden on carers and healthcare practitioners and providing reassurance by by always knowing where a person with dementia currently is, and giving the patients themselves greater freedom.Working with healthcare professionals, students of Healthcare technology students at the Avans University of Applied Sciences are testing healthcare innovations in the “House of Tomorrow (dutch site) “. Students can gain practical experience using innovative technologies such as the IoT at the Care Innovation Center in West Brabant.
IoT in research
Cheap, rapidly scalable and making maximum use of the advantages of a LoRa network”. This is how Matthijs Kooijman described the technical foundations of the Measure your City (Meet je Stad) project. It is an ideal example of how the IoT can be used for research purposes, where, as Dekkers previously indicated, researchers are able to set up the entire network themselves.
Dutch presentation: Measure your City! Amersfoort measures climate change (Matthijs Kooijman and Diana Wildschut, Amersfoort University)
IoT as an ecosystem
Building a cheap and scalable network from the ground up is exactly where The Things Network (TTN) started in 2015: a design working through IoT communities that originated in the Netherlands, and now either has been or will be implemented in around 150 cities worldwide. It’s built from the ground up, unlike national networks such as LoRa, set up by the operator KPN. Giezeman (TTN) argued that the IoT is mostly about good business cases, about ecosystems of applications that offer genuine added value. The network, which is TTN’s main area of activity, is just one component.
From the very start of the TTN network in Amsterdam in late 2015, the aim was to host services on the network that would prove its added value. For example, a service for monitoring whether or not boats on the Amsterdam canals are taking on water.
With the “smart mousetrap” example, Giezeman illustrated the need to create an ecosystem of IoT applications. The smart mousetrap tells facility management providers which mousetraps contain a mouse. This makes the process of routing to empty the mousetraps more efficient.The data can also be used to determine the location of the mouse infestation and stop it at its source. Facility managers can also expand their services, for example, by monitoring when First Aid kits have been used and therefore need to be replaced in an office or factory. This results in an integrated ecosystem of applications.
Presentation: The Things Network (Wienke Giezeman, The Things Network)
IoT from project to commercial application
IoT is as diverse as it is old, explained Paul Havinga (professor of Pervasive Systems, University of Twente). A conference about Smart Dust in 1999 was his first introduction to what was then known as Wireless Connected Sensor Technology (WCN). From 2001 onwards, the University of Twente started its first projects to develop IoT applications that could eventually be implemented commercially. Initially, these were mesh networks, with sensors mainly communicating with each other. Later, they also communicated via mobile or specially built IoT networks.
Just like Giezeman, Havinga emphasised that IoT applications must be part of a broader ecosystem and must have a clear business case. But the disruptive impact can be large, if existing processes are improved and entirely new processes are facilitated.
Project and practical application
Projects from 15 years of IoT research were discussed. Here is a small selection:
- The Eyes project in collaboration with Nedap: monitoring the health of cattle on a farm, finding a specific cow in a huge cattle shed. This led at a later date to the smart parking system for the UT campus.
- COBIS project with SAP (Collaborative Business Items): acquiring real-time information, for example, of how chemical components that may not be placed together are being stored. Checking the implementation of general business rules in order to prevent a fireworks disaster such as the one in Enschede.
- Aware Project 2007-2011: Preventing incidents by identifying potential sources of fire during a heat wave, in collaboration with the fire brigade and others. It used a combination of drones, sensors dropped by drones and “smart clothing” worn by firefighters.
In other projects, self-sufficient sensors (solar or wind power) were used to determine measurements during the construction of a bridge and remained in the bridge after construction to monitor any potential structural problems. Another measured the temperature of train wheel bearings. This had two aims: to prevent trains from flying off the tracks and to make maintenance more efficient: it is carried out as needed rather than regularly.
A high profile project using a combination of LoRa and a mesh network is currently underway in a safari park in South Africa. Among other things, animals fitted with sensors indicate through abnormal movements when poachers are in the area.
According to Havinga, anything can be used for measuring, including mobile phones which can, for example, act as sensors to measure vibrations and identify potholes on roads. Since almost everyone owns a smartphone these days, they have substantial processing capacity and mobile networks provide coverage everywhere, they are ideal sensors.
Presentation: Celebrating 15 years of IoT in Twente(Paul Havinga, Twente University)
What can SURFnet do?
At the start of the seminar, Maurice van den Akker and Paul Dekkers mentioned that SURFnet wants to help institutions get to grips with the technology and practical applications of that wonderful but rather broad concept of IoT. One way SURFnet is doing this is in collaboration with TTN. Together they are investigating the possibilities of developing open courseware for education.
To conclude the seminar, an online survey was carried out in real-time to identify the opportunities and needs of the education and research institutions, and how SURFnet can support them in these. This survey revealed a clear need for an improved understanding of IoT ecosystems rather than of the individual elements.
For a good half of the respondents, it was the applications that attracted the highest level of interest. This could be something simple such as placing sensors in chairs to give students a real-time view of available seating in a lecture hall. But those present saw more opportunities for implementing IoT in research, and less directly in education. This obviously needs some more in-depth research.The response to the network portion of the seminar was more decisive. Three quarters of the respondents saw the value of coming together to create a LoRa network with national coverage, of whom two thirds also indicated that their own institution would be prepared to install a LoRa gateway.SURFnet can expect a warm welcome for helping institutions create sub-communities around the various IoT subject areas. If SURFnet takes the initiative, more than 80 percent of the respondents said they wanted to make an active contribution.
In short: IoT inspires great interest, but even more questions.