Internet of (all those) Things – Hackathon applying wireless sensors
In DNV GL we talk a lot about being “data smart” these days and we are soon launching our industry data platform Veracity, which is closely related to technologies like big data, internet of things, machine learning and more. But what do these words actually mean? In this blog-post, we will try to simplify it a bit and mainly focus on the terminology “Internet of Things” or “IoT” which is a common abbreviation. We will also provide a very specific and easy-to-understand example from a hackathon with wireless sensors we did recently. The “things” within IoT are mainly about sensors measuring or tracking various data. You are surrounded by them in your consumer-life and probably as a professional working in any type of domain or market. A recent Gartner-forecast says that we will have more than 8 billion “things” connected during 2017. That is an increase of 31% since 2016. That’s a lot!
If you look into Google Trends and investigate the search frequency on words like IoT, big data and machine learning, you will see an interesting trend for the past few years. The blue curve represents IoT, the red curve big data and the yellow curve machine learning. As you can see, big data started to increase towards the end of 2012, while IoT came a bit later in the middle of 2014. Anyhow, it is a significant increase in searches for these keywords for the past 3-4 years, so obviously there is a lot of interest and traction around it. A part of the history is the momentum on cloud computing for the past years. It would not be possible to benefit that much from IoT without the enormous amount of compute and storage capacity you have within cloud computing.
When we refer to the Internet of Things, it can be challenging to understand what a “thing” actually is. Some people refer to it as a kind of swarm of sensors and internet-connected devices surrounding us and making us able to make smart decisions.
But, let’s try to be a bit more specific. A “thing” (in IoT context) could, for instance, be anything of the following items:
- A significant asset, i.e. a ship, offshore structure, wind turbine, car or whatever asset
- A mobile phone
- A pure sensor
- …and much more (you can even find IoT-based or internet connected umbrellas! However, I doubt the necessity of such devices)
Even though there are many opportunities within “connected” ships or cars, it is a more complicated thing to explain. It is also more like an IoT ecosystem. In other words, all the components that together can establish new digital business models by connecting IoT devices and including dashboards, networks, gateways, analytics, data storage and security.
As with any technology, you sooner or later have to deep-dive into it to really understand it. If you have not done it already, you should consider purchasing some simple IoT devices that can actually give you a slightly more efficient life if you use them correctly. As mentioned above, we also decided to go ahead with a hackathon, where we played around with some really advanced industrial type of sensors and created a simple facility monitoring system. Below you will find two examples, one simple and one slightly more advanced.
Simple example – IoT “things” you can apply in your daily life
In your private life, you might spend quite some time running around and trying to find the car keys…or what? Where are they? Is it in a bag, in the garage or in some jacket? To deal with this problem you may purchase a bunch of TrackR-devices as one of many options. This is a small device that can connect to your smartphone via a Bluetooth connection. Once your car key (or any other thing the TrackR is hooked to) is within range, it will send its position. If you are not sure where your keys are located, you can use the smartphone app to look for the latest known position and fire an alarm so that your TrackR-device makes some noise. Since Bluetooth has a limited range, several TrackR devices can set up a mesh network to interconnect more devices.
Advanced example – Hackathon with industrial wireless sensors and cloud connection
Earlier in 2017, we got curious about the Norwegian IoT startup Disruptive Technologies (DT) and their sensors during an event organized by Webstep in Oslo. Disruptive Technologies recently became the regional winner of the Nordic Startup Award 2017 within the IoT-segment, so they have a very interesting technology and roadmap. However, there are many new IoT companies coming up these days and most of them have their unique capabilities or business models. Anyhow, it was more efficient for us to establish contact with some local players providing advanced technology and know-how and see where it could take us within a short 3-day hackathon.
As always, it is good to have some kind of goal or “theme” for a hackathon. Our goal was to develop a simple facility management solution targetting an innovation room that we’re designing in our new office landscape. The current focus for our hackathon-partners (DT and Webstep) is to provide a very secure, scalable and robust platform first (“sensor-as-a-service”), rather than many sensor types at once. The pilot-kit (SDK) we used in the hackathon had 3 types of sensors: on/off (like a switch), temperature and proximity. One of the interesting things we learned was that the battery life for a sensor would be up to 10-15 years, but of course depending on how often it sends signals (in our case it was per minute or on-demand). Another impressive thing was the signal range that we measured to 30-40 meters indoor and several hundred meters outdoor, all depending on conditions and obstacles. In addition, the sensor kit provided great usability. It was a “plug and play” experience and everything worked smoothly from the very start. The sensors are sending wireless and encrypted signals to a “cloud connector” which again streams it up to a cloud provider (Amazon, Microsoft, Google and more).
The picture on the left side shows the “cloud connector” which can be connected to your cable network, Wifi or mobile data.
The picture to the right shows a proximity sensor inside a door. It is relatively small, so it would fit into a small space and send signals when the door.
The picture below is our “connected” reading lamp using a Particle Photon device to hook it up with a relay and a power source.
The sensors are sending frequent (encrypted) signals to the cloud connector and then into the cloud. In our case, we used a combination of Google Cloud and Microsoft Azure.
The illustration below shows the high level “architecture” of our IoT-solution. After all the sensor data has been streamed through the cloud connector, the Google/Azure cloud, it ends up in a web application containing a 3D-model of the innovation room. We have a dashboard there with charts and notification areas responding to changes in the sensor data.
Our simple facility management dashboard could show the following information:
- Facility insights, like room temperature, fridge temperature and so on
- “Seat occupied” alarm and “seat counter” (using proximity sensors)
- Door open/closed indicator (using proximity sensors)
- Automatic “reading lamp on” when people sit down on a chair (using Particle Photon and sensors)
- Alarm when fire extinguisher is removed (using proximity sensors)
- Alarm when other important objects are removed, e.g. our valuable Microsoft HoloLens (using proximity sensors)
- Customer survey, giving the audience the possibility to rate their experience (using 3 on/off sensors)
Although we spent 3 days on our hackathon (with 6 developers from DNV GL and Webstep), we achieved a lot and got our “hands dirty” with some really exciting technologies. We got a far better understanding about being data smart and the business potential with sensors, connectivity, big data and machine learning. You are far better equipped to deal with “things” like ships, offshore structures or process plants at a later stage once you understand the potential with some emerging technologies.