Research Summary - Convergence of Blockchain, IoT, and AI

Hey Yen, thanks for the comment and welcome to SCRF! That’s another great convergence idea that you bring up: using AI to recommend different blockchains for different IoT devices.
In terms of theoretical vs. applied research, this paper is much more on the applied research end of the spectrum. It is true that this paper is highly concept-based and takes on more the form of an idea generator than of a finalized implementation. Nonetheless, even the most sophisticated of technologies start off as just an idea.

2 Likes

Great and easy to read summary!

The convergence of Blockchain, AI and IoT technologies to create systems and applications may be inevitable and will, if/when it happens, turn out to be quite beneficial to most people. When combined, these technologies can boost new business models and disrupt the markets as we have them now.

Artificial Intelligence is able to analyse massive data to extract insightful trends and information from those data. Blockchain will also solve most of the security issues that arise from IoT and AI. Blockchain could also provide better privacy protection and promote ease of payment.

However, to be realistic, managing these technologies singly involves high costs and risks, nonetheless when they are combined. Also, there isn’t enough energy to power systems that will arise from the combination of these technologies. Currently, gas prices are rising exponentially and may continue to rise if there is no deterrent, which in turn, means Blockchain may prove to be far more expensive and difficult to afford. There is also the issue of Bandwidth, with the combination of these technologies, we will need faster and more efficient internet. 5G, I believe, is only the beginning of the sort of Telecoms tech that will be needed. Perhaps, as stated in the key takeaway, by the time these combined technologies are ready for use, these challenges would have been largely surmounted.

3 Likes

I found one of the comments in this paper very insightful: AI is an automation tool for decision making and it can optimize the way we store data as well as decide on which data we want to store. This characteristic of AI can benefit IoT in controlling facilities in a system. The data that is produced by IoT needs a large and secure medium for storage. This is where blockchain comes in.

Recently I read an article on risks of applying blockchain in IoT (I linked it down below). I extracted and summarized some of the most interesting and important aspects of this source.

Challenges and risks of using blockchain in IoT:

  1. Scalability

Blockchain makes use of decentralized networks in its operation. However, as the size of the ledger increases, the network might become centralized instead. A record management system is required to keep the centralized network in check.

  1. Processing power and time

An IoT system consists of many devices with different levels of processing power. Not all devices can run the same encryption algorithms. In addition, some parts of the IoT system might experience lag.

  1. Storage

Due to the decentralized network in blockchain, the ledger record should be stored on the nodes. The amount of storage needed increases over time, and might even surpass the capabilities of the facilities in an IoT system.

  1. Vendor

As blockchain is a novel technology in this era, human resources with strong technical background and skills are still scarce. To overcome this issue, many organizations and enterprises should hire specialized vendors to set up blockchain solutions. However, it is often hard to find vendors with high credibility.

  1. Security

Despite blockchain’s high security, it is still prone to malicious activity from other parties. This can happen especially on a public blockchain, where many users have a private key to “sign” on transactions as most systems lack multi-factor authentication. In addition, losing the account’s private key means losing the account’s data as well.

  1. Legal issues

The laws governing blockchain and its usage differ among countries. Many businesses prefer to avoid blockchain altogether due to the risks of legal issues.

All in all, there are many elements to consider when applying blockchain in IoT, and I think that among the four convergence cases you give (AIoT, IoT Blockchain, Blockchain AI, and AIoT Blockchain), IoT Blockchain and AIoT are much more studied, yet there are still many unknowns ahead even for these convergence cases.

8 Likes

The Pay as You Go Business Model

This is article is a great exploration of what’s possible, and I absolutely welcome such explorations. At the same time, I’m reminded of a movie line in which a character lamented the move from science to business without differentiating strongly between coulds and shoulds.

The practical consequences of some of the scenarios proposed are interesting. For example, the pay-per use model has been long discussed among practitioners in the IoT space. However, the day-to-day consequences of living in a world in which your devices may cease to function due to not putting enough quarters in the machine have prevented that model from taking shape, at least in the B2C space.

If a consumer has a smart lock that only functions when the lock bill has been paid, the consumer could find herself locked outside or inside of her home. The authors also provide a B2G case of a privatized street lamp, a common public good. Hooking a safety device up to only function when it’s been paid presents a special kind of public danger. Even in a B2B case, tokenization of multiple collaborating devices could lead to stalled work not because of anything wrong with the device itself but perhaps some particular issue with the blockchain of that specific device, adding an additional layer of complexity–one the owner or some technician may be incapable of addressing.

The case for AI assisting the management of IoT is easy. The case for blockchain is also exciting, but in order for the vision laid out in this article to come to fruition, it may take a clearly communicated value proposition that convinces users that individual bills for every device they have are a benefit rather than a complex tapestry of bills to stress over. If an edge device has the potential to quit on a user (camper, explorer, hunter, etc.) in a far off area because of triggered condition in a smart contract, even though the device is otherwise well-formed and functioning, this could be viewed as a tremendous reduction in value. So, while the data utility and business value is clear, I’m curious if anyone has come across studies on how these implied and proposed blockchain enabled pay-per use business models are structured, perhaps with some hints as to their practicality or viability.
ianMalcolm

2 Likes

… and almost one year later, Seth returns :man_facepalming:

You bring up a great point about energy consumption for AIoT devices that power blockchain, and there is some (albeit limited) discussion of this in two of my AIoT Blockchain papers: MLAV and Parameter Space Exploration. In a nutshell, consensus algorithms and verification methods both have huge implications on power consumption for blockchain systems. In MLAV, we replace traditional signature verification with a novel aggregate verification method, and in Parameter Space Exploration, we find that Crash Fault Tolerance (CFT) is an inherently energy-efficient consensus method - a good choice if there is no risk of malicious nodes in the blockchain system.

Needless to say, lots has happened ever since your comment was posted: further increased gas prices, major dips in the price of cryptocurrencies across the board (their prices were getting ready to peak when you made that comment), and huge breakthroughs in AI such as we see with DALL-E 2, quantum AI platforms developed by NVIDIA and Google, and even talks of GPT-4 development.

I’ve also been hearing talks about 6G development recently, but I haven’t done any research on that yet… do you happen to be familiar in that area?

2 Likes

Thank you for this information @Gearlad , I’ll like to bring out my claim point of see on this that IoT-blockchain frameworks fulfills the requirement for straightforwardness, responsibility, secure exchanges, lower costs;This is when blockchain comes into play, not as it were since of the adaptability it gives, but moreover most critically its unchanging nature feature.
Looking into this, utilizing the information, manufactured insights (AI) can progress the developing and offering forms with the parameters put in their model.
Furthermore, each gadget ought to be prepared with the capacity to communicate with each other and make the suitable choice in time, which is the substance of AIOT. Therefore, choices made by AIoT gadgets do have to be reasonable.

Reference

I read about IoT and AI from this source and I feel I should refer anyone interested to it.

Blockchain, IoT and AI — A Perfect Fit | by Philipp Sandner | Medium:
https://www.google.com/url?sa=t&source=web&rct=j&url=https://philippsandner.medium.com/blockchain-iot-and-ai-a-perfect-fit-c863c0761b6%23:~:text%3DA%2520possible%2520connection%2520between%2520these,full%2520potential%2520if%2520applied%2520combined.&ved=2ahUKEwiUn9TIv7X5AhUQt6QKHQvVA_IQFnoECAsQBQ&usg=AOvVaw0rC9yvF5C3FYpD9H3bJx1w

4 Likes

Welcome back Seth.

Haha, yes, a lot has changed since my comment. I am not very familiar with 6G, however, I have heard of exciting developments in that area. NEC Corporation is collaborating with some other companies to develop a distributed Multi Input Multi Output (MIMO) technology which will expectedly increase the stability of communication in high-frequency bands among others. Another interesting intersection is that some researchers believe that collaborative artificial intelligence will be a major driving force in the success of 6G.

1 Like

@Gearlad
Nice summary with lots of information and contributions to digest. Blockchain, AI and IOT have been incorporated in every technical innovations and these threes have come to make our lives better and technology have gotten better with them. Right now, technological evolution can not do without them. Every aspect of technology is incorporating them.
AI has been incorporated into smart cities to analyze data in real-time and monitor activities.
Due to its adoption of blockchain-based innovations, Dubai is one of the most technologically advanced cities and will soon be recognized as the first blockchain-based metropolis.
Additionally, by utilizing IoT devices and maximizing the usage of electricity, water, and other resources, smart cities can lower resource consumption and traffic congestion.
IoT and blockchain work together flawlessly to provide safe data and advance commercial transactions.
The integration of blockchain, AI, and IoT will increase each technology’s advantages while reducing its hazards.
IoT is linked to devices where data can be stolen, fraudulently obtained, or hacked.
Blockchain can assist to maintain transparency and security while AI can help protect it from malware and hackers.
Combining all three technologies will increase their potency and fortify the technology’s weaker points in order to make the future better for everyone.

2 Likes