IoT and AI revolution: Maximizing connectivity, while minimizing TCO

AI and IoT are very powerful technologies. AIoT is now happening. Big impact will happen across automotive and connected cars.

Pradeep Chakraborty
Updated On
New Update
AI and IoT

AI and IoT

Next, there was another panel discussion around the IoT and AI revolution: Maximizing connectivity, while minimizing TCO. It was moderated by Minu Sirsalewala, Executive Editor, Special Projects, CMIL.


AI and IoT focus

Bhushan Sethi, VP, IoT Product Solutions & Onboarding, Vodafone Idea, said that a lot of discussion has been around AI/ML, deep learning, neural networks, etc. AI and IoT are very powerful technologies. AIoT is now happening. Big impact will happen across automotive and connected cars. Facebook has 30 million lines of codes. Connected car has over 100 million lines of codes. There are different levels, and the maximum level is to be autonomous. There are going to be over 200 IoT sensors in a connected car. They will talk via IoT to the AI system via the edge. 

Another one is Industry 4.0! Today, manufacturing contributes 15% to the GDP. It has be over 25%. Cloud and IoT will help build better designs. Third area is around medical. We will see personalized medical. Fourth, we are seeing smart energy. Power will flow in a dynamic manner. There are some concerns to be looked at. Privacy, cyber security, data algorithms, security, etc., are areas.

Anand Bhandari, Vertical Head, IoT, Reliance Jio, said every domain needs IoT and AI. All domains will get transformed. AI is utilizing data. Right data, at right time, that is most reliable (RRR) will be most important. Data is the oil, and AI is the creature dwelling on it. The ecosystem of edge devices to the platform has to be owned by someone. We also need to have energy sustainability. AI will help us get to net zero emissions.


Sethi added that we can collaborate with AIoT, and it is coming together. Magic will be coming from developers. Both technologies are partnership driven. They will enable use cases around Industry 4.0, automotive, healthcare, etc. AI will also enable IoT. Billions of devices are placed remotely, and we are sending data. We can now troubleshoot using AI/ML over IoT. AI/ML will apply same logic for devices say, 10 days later.

Scaling costs of networks

Puneet Chopra, Telecom CTO, HPE India, added that we are seeing new technologies are bringing in top-line benefits. They will grow up and scale. Core networks were vertical silos. As they advanced, they were cloudified. Things were again silos. Now, we are having horizontal cloud platforms. We are having SD-WAN, MPLS. etc., as-a-service. We are looking at scaling costs of networks in future. We are having the hypervisor to become virtualized. Today, containers are there, and are easy to compose, manoeuvre, and orchestrate. 

HP is doing work for virtualized architectures. We also have cloudified architectures getting converted to cloud-native architectures. We need to tell orchestration manager what we need to achieve. DSTN can build different models that separates layers. We are able to have assurance, as well, with contribution from AI/ML. As it improves over time, we can predict networks from degrading, network faults, etc. 


Bhandari added you can detect any fault, and send it across. We can also collect data from multiple devices, and send only the necessary data. SCADA has been working for a while now. Local intelligence can be used for real-time decision making. We can also suggest improvements in the processes. 

We also have security of hardware, connectivity, etc. Attackers are also evolving. We need to see how network patterns are behaving. Algorithms can detect folks logging into the system. They can add third authentication, if needed. AI algorithms can help protect the data. 

Bhushan said data will stay at multiple places. If it is V2X, it cannot go to other places. Data security is equally important at every place. DoT has recommended some stages. Security aspects need to be taken care of very well. 

Intent-driven orchestration

Chopra added that we need intent-driven orchestration. The machine should apply its own mind to achieve an intent. We need to have proper load balance. Engineers can spend more time in designing and solving problems. 

Bhushan noted that partnership is very important for the ecosystem. Most IoT apps are silos in nature today. Devices are also probably in similar state. Standard has to be made compulsory for IoT side. Bhandari noted that collaboration is key, and for taking industry opinions.

ai iot