|
Listen to this story
|
Today, in the modern Internet of Things (IoT) age, a record amount of information created by linked devices needs to be gathered and evaluated. As a result, massive amounts of data are generated in real time, necessitating the use of AI systems to interpret this volume. Edge AI refers to the installation of AI software on hardware throughout the real world. Every sector in business is trying to boost automation to enhance workflow, productivity, and security.
With the help of developments in edge AI, robots and gadgets can now function with the “intelligence” of human cognition wherever they may be required to. An industry that is resilient, human-centred, and sustainable is the primary focus of Industry 5.0, which is a complement to Industry 4.0.
So, how do we orchestrate this transition into Industry 5.0 where the emphasis is on interaction between humans and machines?
The genesis of Industry 5.0 is that this next generation of manufacturing, or producing goods, has to go beyond merely making something consumer-focused because consumers always want something faster, better and cheaper. 5.0 is leveraging technology to not only be consumer-focused, but also create a company that is worker-focused, sustainable, and that is also focused on the country that it is operating in. Gartner’s Magic Quadrant, even today, put Edge AI at the peak of inflated expectations, which means that it is not mature yet.
- Purnesh Gali, Head of Analytics at Hindalco
The Roundtable Session was moderated by Purnesh Gali, Head of Analytics at Hindalco with panellists Biswajit Biswas, Chief Data Scientist at Tata Elxsi, Muthumari S, Head of Data Science at Brillio, Pradeep Sukumaran, VP, AI & Cloud SW at Ignitarium, and Mathangi Sri, Chief Data Officer at Yubi surrounding the agenda of Edge AI and its transition into Industry 5.0.
Edge AI and its definitions
Edge AI is nothing but applying our AI-based algorithms, which are processed within the data collection, hardware itself. Edge AI is used when you want real time inferencing and can be a use case in banking or retail and predictive maintenance or when we have to do any.
- Muthumari S, Head of Data and Analytics at Brillio
The transition from industry 4.0 to 5.0 was all about IoT as the main thing that happened that era and now, we are looking at adding intelligence into this and have a new term called as AIot. Edge AI basically means acquire data, do the processing at the edge and send the inference over to the next layer, which could be an Edge Gateway or the cloud. What makes it happen is the compute.
- Pradeep Sukumaran, VP, AI & Cloud SW at Ignitarium
Edge AI is where we are performing AI at the source of Data. Edge is still doing inferencing and filtering the data that you need. One of the reasons why Edge AI is attractive is because we may send every data to the cloud and there is nothing in between. What Edge AI can do is send the signals which need to be further processed and further looked at. So a tiny model, tiny ml, or AI model, which is designed not to do an inference but to shape and filter your data.
- Biswajit Biswas, Chief Data Scientist at Tata Elxsi
Edge use cases are very important. We have not started exploring, but one of the potential areas could be on image processing, or when we are scanning a particular place, whether we want to use Edge AI there in order to make a decision at that point in time without contacting the server.
- Mathangi Sri, Chief Data Officer at Yubi
Edge AI since time
Edge AI has been there but its processing power has been different. You can’t just integrate it with cloud, but to some extent, within the machine, you can configure a lot of things manually, physically, that requires a very different kind of skill—maybe it was not called AI at that point in time, like the definition that we have right now is a little bit different. Intelligence at the edge has always existed in some shape and form, it is just that the maturity is significantly different now.
- Muthumari S, Head of Data Science at Brillio
Is Edge AI Manufacturing or Computer vision?
When Industry for edge comes to the FinTech world, there is certain sensitive information that you may want to communicate back to the cloud and if you want to make it for that particular user, you want to make decisions based on certain signals that you get from the phone, and the data that is there in the phone—that is also where edge can be used. Those are two different things other than the computer vision, but these are the core user behaviour kind of use cases wherever you want a user behaviour case. Another is a use case where you don’t want your data to get mixed up with other users’ data and pass it back. It’s very sensitive data and hence you have InfoSec issues, which you want to resolve at the phone itself. Third use case also, which is about monitoring logs on the mobile or anything around that, also can be a very good big edge use case where you are equaling a huge number of logs and do not want to pass it and load your service and make decisions on the go.
- Mathangi Sri, Chief Data Officer at Yubi
How to decide edge versus non-edge use cases?
It depends on the industry that you’re dealing with, the application, the customers, because there is a cost benefit analysis that needs to be figured out. But, a rough thumb rule that I can say here is that it depends on the amount of data that you have to deal with. If you’re looking at an autonomous forklift, you would want to do the processing at the edge. But if you’re talking about a lighter swarm of robots, you might want to make it low-power battery operated, and hence not have so much compute. So, you would then have some compute there and the rest happening at the cloud orchestration. I think that’s a case-to-case basis, you can decide whether it’s cloud or edge.
- Pradeep Sukumaran, VP, AI & Cloud SW at Ignitarium
Connectivity in Edge AI
Connectivity is very central to edge computing. I think about how you connect these edge devices for collecting data: the sensors collecting data to the device and device then connecting and doing numerous uplink and downlink communications with its server. Connectivity is central. But, there’s one single place we all get stuck on—how to get the data out from the machine. The answer is that it is still being solved and under escalations.
- Biswajit Biswas, Chief Data Scientist at Tata Elxsi
Edge AI can be used despite connectivity. So, I was talking about governance. Wherever you cannot have user specific data being stored in the server, it is a very good way to resolve that with the help of governance. For these reasons, you may have to do Edge.
- Muthumari S, Head of Data Science at Brillio
Need Edge AI around 5G?
5G will give you backhaul connectivity—the last mile connectivity or what they call ‘Fog Computing’. That has to depend on the local short range wireless or the LoRA protocols because those protocols are designed to carry your data with appropriate latency. 5G is good. But, it won’t take care of end-to-end immediately. You still need to have your Fog Computing layer taken care of very well.
- Biswajit Biswas, Chief Data Scientist at Tata Elxsi
Using Edge AI as a strategic advantage
Edge AI can be an advantage if firstly, we think of edge computing as one of the architect’s resolutions. Secondly, thinking of edge computing as one of the architect’s resolutions. These two things, if that could change, it could change the nature of how we compute today.
- Mathangi Sri, Chief Data Officer at Yubi
More businesses and developers are seeing the value of implementing Edge technology to deliver faster, more effective service while increasing their profit margin as customers spend more time on their mobile devices. As a result, there will be a wide range of new opportunities for enterprise-level, AI-based services, user comfort, and satisfaction.
Edge AI is already in use and will be the cutting edge of future technology. Today, there is a race to the top among rivals because there is an increasing requirement for data processing at the edge as a result of the growth in the volume of data generated by devices. “The Fifth Industrial Revolution is evolving from a concentration on the digital experience to one where humans are back in charge. The results will combine the skill and speed of automation with humans’ critical and creative thinking,” says Dan Gamota, VP of manufacturing, technology, and innovation at Jabil.
This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.
