The shift to the edge
An interview with Aaron Ding on the promise of edge computing
As the Internet of Things (IoT) is growing, so does the attention to edge computing. ‘We’re currently shifting to this new computing paradigm’, explains TU Delft Tenure-Track Assistant Professor Aaron Ding. Already in the foreday of edge computing, Ding’s research focuses on IoT and the power of distributed computing resources that are deployed at the network edge.
What was the old paradigm?
‘To understand edge computing, one needs to understand the old paradigm: cloud computing. Imagine asking a smart speaker to play a song for you. This speech file will be uploaded to the cloud. In other words, through the Internet the speech file will be sent to a server somewhere else. That server then analyses your request using algorithms and sends back the song so can listen to it in your living room.’
Why do we need edge computing?
‘There are two main concerns about cloud computing. The first is the issue of privacy. More sensitive information can be collected nowadays by different sensors and all that information goes into the cloud. Since many of these servers are not located in Europe, cloud providers do not necessarily have to comply with the European privacy laws.
The second issue is a data traffic issue. Nowadays, we have too much data generated by different devices. Think about the future in which we will have thousands of smart cars, and they will be generating terabytes and terabytes of data. If all that data has to fly through the Internet to the cloud - many in the US - there is too much data that has to travel back and forth. You can compare this problem with highways: if there are too many cars, the roads will be completely blocked.’
What is edge computing and how does it tackle these concerns?
‘Edge computing is the new generation of cloud computing. We bring the cloud computing from the fay away data centre closer to the user. Think of a smart speaker that collects different sensing information from the environment. Instead of sending this information to the cloud, they process and analyse this data on the spot. By having the computing and processing nearby – on the spot – the user has control. In other words, we are changing from do things centralised to doing things decentralised. Thereby, we avoid giving potentially one single central party control over everything. Also, edge computing is trying to avoid data redundancy. If we can already process half of the data nearby, then we reduce the potential data traffic flow by a half.’
Why do IoT and edge computing interest you?
‘IoT is about increasing connection. It makes it possible to capture all our movements and all our actions and translate that into knowledge. Every day, we do a lot of things and every day, we actually want to improve certain things like healthcare and well-being. All information captured by IoT can help with that. The whole magic of IoT is that it can bring our society to the next level, by making it more intelligent, more sustainable and healthier. However, we have to find a correct way of implementing that in our lives: it has to be sustainable and privacy-preserving.’
How far is the edge computing technology and what are the current challenges?
‘Edge-computing is very rapidly developing nowadays. Several major companies are producing chips to do edge computing, like Google, Amazon and Microsoft. They all have the hard- and software platforms to support this. It is not fully entering the user market yet, but for some use-cases in the industry it is already used.
One of the problems in this field is that we tend to think that engineering is the solution for most of the problems. But the field is getting more mature: we have more human factors and more business concerns. Some of the problems are so hard, that is impossible to solve just by pure technical mechanisms. That is why, in addition to developing our own edge-network, we’re touching upon other social, socio-technical and economical disciplines during our research.’
Aaron Ding and his team are involved in the Delft on Internet of Things (Do IoT) Fieldlab. We offer an open platform focused on the development and use of Internet of Things and 5G. Researchers, companies, government and students can develop and test IoT applications based on the latest wireless communication technology in collaboration with the fieldlab. Together, we accelerate innovation in mobility, industry, logistics, agro, health and safety. Follow us online on https://www.tudelft.nl/LinkedIn or Twitter!