With Aime, we specialize in crafting smart software powered by Artificial Intelligence & Big Data. To make a long story short, that’s about building software which supports your business processes in a data driven way. Mobile apps and web appliations which can be used in the office and in the field, using modern technologies and – wherever desired – the cloud. That’s what we do on a daily basis.
In doing so, we use data that is already stored on your shelf. For example, we recently synthesized chemical data flows in a successful way for Nouryon. As a result, a training set for machine learning models is now being created automatically.
It is only one of the examples which shows what you can do with sensor data. But when we’re talking about sensor technology, we often hear something about the Internet of Things, also known as IoT. We wondered: what is this internet, exactly? Is it different from the internet we know? The answers to those questions are covered in this blog. But since we build software that is powered by AI and Big Data, we also link the IoT to the machine learning technique of AI. In the end, we will discover the nice way in which the IoT and AI can be combined!
Cover picture: blickpixel, Pixabay license.
What is the Internet of Things?
Connect the world. That’s the slogan of the internet that we know: devices which can be controlled by human beings. Devices that are networked and which allow us to communicate with each other. By using computers, laptops, tables and phones we communicate through websites, through apps like Whatsapp, but also through more traditional channels such as email and IRC.
Now suppose that you are a logistics company which has currently many trucks under way for delivery. Specifically, we’re covering cooled transport. As a logistics company, you’re in big trouble if the temperature rises above a certain value. Additionally, your customers requests real time insight into the location and the temperature of their delivery. This is the perfect entry scenario for the Internet of Things.
In this IoT, the scenario that we discussed above – devices that are connected for humans to communicate – becomes a minority between all other devices. Rather, objects will become the majority of devices connected through the internet. Very normal objects, in fact, as you can see in the image: flowers that are equipped with sensors and share insight into the status of plant feeding, for example. Or a smart home of which the sensors are linked to the internet. We can see a new type of internet arising: one that can be used to share information about things!
As you have probably assumed, sensors play a critical role in the Internet of Things. It’s those sensors that enrich regular objects and measure data necessary for your use case. As we saw in our blog about the combination workshop IoT + AI, startup Iotify is specialized in IoT sensor technology. They have for example created an OmniSensor, which is a device to which various sensors can be connected. This grants you extreme flexibility with regards to the measurements you need. Their OmniSensor can be combined with a GPS sensor and communicates over a relatively inexpensive and wireless network connection – LoRa – with the internet. In terms of the logistics company example: here’s your realtime insights!
Artificial Intelligence and the IoT
Now, pretty quickly, another question emerges: nice, this IoT, but how are Artificial Intelligence (and machine learning and deep learning) related to this network? What is the rol for Big Data?
Smart question. That role is very much real these days.
As we saw during the combiend workshop: the IoT sensors generate Big Data, which can be harnessed by Aime to generate insights with respect to your original measurement. According to Iotify, it makes Aime some kind of gold smith – being capable of retrieving insights from measurement.
Iotify mining the gold, Aime being the gold smith – that’s a pretty nice naming convention.
Besides being funny, it also clearly shows the interplay between AI techniques and the IoT. Where the internet-connected sensors give realtime insight in the current situation, they can also be used to generate a very rich AI training set in a relatively short time. This training set can be used to train AI models which can discover patterns and provide you with predictive power for new data. Very real examples are:
- Predicting the level of oxygen in offices based on time of day and the level of oxygen during the previous 30 minutes;
- Identifying the odds that someone is still present in their offices at a certain time, which allows you to make an assessment whether it is worthwhile to visit the colleague – or to drop by tomorrow;
- Predicting machine failure, such as aircraft engines, based on real-time measurements. This allows them to incur less damage. It also allows maintenance crews to plan maintenance in a smarter way.
- To make parts of roads smart and to allow them to indicate that they are broken.
And so on!
Aime is located at Business Innovation Campus de Gasfabriek in Deventer. On the innovation campus, various startup companies work on technology innovations. New technologies, such as sensors and machine learnings, play a key role in those developments. Nevertheless, it is not about the technologies themselves, but rather, about what you can achieve by adopting them. That is: to design your business processes in an entirely different way, with respect for human and working culture, and reaching new customers.
Did our specialist toolbox trigger some kind of interest? Obviously, we would be more than happy to get in touch. We serve delicious ☕! Get in touch.
- Date - 1 February 2019