Until now we have seen 3 industrial revolutions. The 1st revolution used steam power to further production while the 2nd used electricity. The 3rd revolution automated production with electronics and IT. Now the world is witnessing the 4th revolution which is in essence a digital revolution. This is revolutionizing the field of technology as we know it and blurring the lines between digital, physical and biological spheres. The speed with which the new technologies are being created cannot be compared with any period in history. The fourth revolution is therefore evolving at a rate which can be said to be exponential. The previous revolutions can be said to have evolved at a linear rate. The 1st industrial revolution started in 1765, 2nd started in 1870, 3rd started in 1969. The fourth industrial revolution is happening right now boosted by the emergence of Internet and digitization as said above. Internet came in 1990s and unlike previous revolutions a host of technologies like Internet of Things (IoT), Cloud Computing have emerged out of internet now itself and not even 30 years have passed. Hence we can see more and more technologies supporting each other and being the reason for the emergence of some other technology. Let’s see how this actually works out.
Seeing it the other way in the computing world until now we can see three platforms. The 1st platform was that of the mainframe with terminals. The 2nd platform was with the introduction of the internet and client/server architecture. The 3rd platform is what we can see trending now where mobile, cloud, big data and social media are dominating the world of technology. IDC refers the 4th platform as Innovative industry solutions. Digitization on the 3rd platform is the trend which we are seeing now where processes, media and intelligence are increasingly being digitized. Cloud is usually used to access that data and make it available in multiple devices. These technologies are materializing and you can see them in glimpses as self-driving cars, home appliances, cognitive systems (IBM Watson, Cortana), robotics(Pepper) and wearable devices. That is where AI and IoT come in. The interactions and operations employed by these technologies should be blazing fast and perfectly secure. This is provided by Blockchain.
DevOps is baseless without the cloud. IoT needs the Cloud to operate efficiently. AI remained only as model up until the advent of big data. The confluence of technologies is just inevitable and often they are very beneficial. Blockchain is the new technology that is taking its head out into the world. Various startups are focused on Blockchain and even governments are realizing its potential. IoT and AI are those technologies which are being implemented in various domains across the world. There has been a huge use of AI technologies in the IoT start-ups in the recent times. Various firms have acquired dozens of smaller firms who are using the AI and IoT technologies in their game. Machine learning based analytics is now being provided by major IoT and cloud vendors who offer a host of AI capabilities.
AI is better able to torture data successfully to garner valuable insights from it and therefore it is playing a good role in IoT. Machine Learning is a sub branch of AI has huge potential to detect the patterns and anomalies in the data that smart sensors generate. This data is related to mainly temperature, vibration, pressure, humidity, air quality, sound. Firms are finding out that Machine Learning may outperform BI tools when analyzing IoT data is the question. Compared to threshold-based monitoring systems the operational predictions using Machine Learning is 20 times earlier and on top of it the accuracy is also greater. If the data operations such as these have to be secure then Blockchain technology can play a vital role.
Blockchain is redefining how trusted transactions ought to be carried out. The internet is itself highly vulnerable and Blockchain is out with the solution to address it. One problem that Blockchain solves of AI and IoT is the security fault lines. Most IoT devices are connected to each other via public networks and it is needless to say how vulnerable public networks really are. Blockchain solves this problem by linear and permanent indexed records which can be created. Globally general public can reference them without censorship. They can also smoothen the commerce process by providing a payment mechanism as well as communication channel. The public is the authority and not any centralized entity as is the case with the banks.
Any kind of hacking and tampering with the data like taking control of device and records is impossible due to the way blocks are stored and guarded in a specific database in the Blockchain system. Every IoT device is a point of vulnerability and the risks are still higher as even AI is involved in making decisions for users. Hence Blockchain can be used to provide a secure, scalable and verifiable platform that has invincible security implementations.
Blockchain solution to centralized IoT systems
Client/Server paradigm is the centralized communication model which is employed typically by the IoT systems. They are expensive because of the huge infrastructure costs needed for centralized cloud systems and large server farms along with a host of networking equipments. Remember we said how Cloud Computing will be the base through which IoT will grow big. The magnitude of the communication that happen will be in the range of billions which will shoot up the costs required to maintain them.
A peer-to-peer communication model can be an effective solution to this centralized model of effecting communication which will significantly reduce costs. But the main problem here is that of security. Blockchain can be used here where it lays out a distributed ledger of transactions shared by nodes in the network instead of a standalone central server. Cryptography can be used to authenticate and identify the partaking nodes and allow them to add transactions to the ledger. These transactions are verified by other nodes and hence no central authority plays a part here.
Instances where all these technologies converge
The focus of AI powered IoT systems are currently on TVs and refrigerators. But it is bound to reflect in vital domains as well. Smart medical devices can be powered by AI to make potentially lifesaving decisions through biometric data.
1. Ethereum Blockchain can make the use of smart contracts very popular owing to its efficiency. The data from the IoT devices can be made to trigger some tasks. Auto insurance is one good example. Assume there are sensors (IoT) installed in a car and in some parts defects appear. The AI will aid the sensor to detect the defects. Consider if Blockchain is implemented then the scenario will change here. As soon as the sensor detects a defect then automatically the insurance money will be credited to the claimant’s account. Before the claimant starts his car and finds out that it is not working the sensor would have detected it and would have transferred the insurance money to the claimant’s account. This is yet to happen but it is a highly probable situation.
We believe two things are main factors for this of insurance being highly relevant in the future. That is we have immense faith in Blockchain technology and smart contracts that we vision it to replace many banking operations. Insurance is another place where it can aptly fit. The second factor is pioneering companies in all domains will always be on heels to leverage the latest emerging technologies. Companies like Berkshire Hathaway, Allianz, State Farm Group are the top insurance players in the world. Blockchain, AI and IoT would have gained immense credibility owing to their potential. The simple reason why they might consider this idea is because the instantaneous settlement of insurance claims through Blockchain will no doubt win the hearts of the public. In all probabilities, these companies will obviously leverage this trusted transaction framework called Blockchain to increase their market share.
But what if the claimants were to purposefully damage the car to claim insurance? This is more probable to happen. Do you know that the homicide investigators in US actually check the beneficiary of an insurance policy and place them as potential suspects in their investigation? That is why to prevent the misuse of the insurance, IoT sensors can be used to detect whether or not the defect has actually happened. But even the best sensors can be manipulated or tampered with. Now you may ask how then such insurance can be provided. That is where AI comes in.
Machine Learning enabled analysis is employed in best of fraud detection systems. This is in stark contrast to the rule based systems where the creators of these systems often make their own narrow views about the detecting system. This is the reason for so many false positives of the customers where the customers are wrongly detected as committing a fraud. The customer will therefore leave the firm and the firm loses the lifelong value generated by such a customer. Machine learning systems are capable to detect the most savvy and intelligent fraud activities at very high speeds, with greater efficiency and with huge scale. They can detect the data from sensor in this case to assess whether or not the defect in the vehicle is genuine or not. False positives will greatly be reduced in this case. Likewise the trio could massively disrupt the auto insurance industry.
2. Progressive is an US car insurance company which is pioneering usage based insurance (UBI) to monitor how its customers drive and is a good example of this. They use an ODB telematics dongle along with Machine Learning which enables them to assess how a driver is performing on each journey. Safer drivers obviously receive lower premiums. They have until now done over 1.7 trillion driving observations and truly the insurance is based on usage rather than the old way of deciding insurance based on where one lives and what kind of car he drives and so on. The firm has partnered with Zubie which is the maker of a device that plugs into a car and can well track how well one is driving. Zubie customers can therefore see how the insurance company Progressive charges them based on the data which Zubie garners.
To make it complete consider if the Blockchain is used to implement the insurance agreement. Then if the car meets an accident then this Blockchain can be invoked to automatically debit from the insurer to the claimant. This process is better than the one where the claimant has to go through various formalities and paperwork to get it done. Not all insurance companies are adept at providing apt services. The time is the critical factor here where it will take considerable time for the insurance company to provide the money to the claimant. In Blockchain smart contract it is instantaneous. This would be a huge thing in the future and surely one insurance company after another will compete with each other to provide better smart contract service to the policy holders.
3. Assume you are planning on a trip to Dubai and have searched extensively on the web for the best hotel at the best price. You receive a notification on your smart watch(IoT) about a proper deal so you go for it. You make payment in Ethereum (Blockchain digital cryptocurrency) to the hotel management. Assume your flight got delayed to Dubai and you are stranded in the airport. You’ll take to Facebook or Twitter and comment on the awful experience you had. The airline has an AI solution that goes through all the comments and tweets you shared in social media and assesses how frustrated you are with the service of the airlines. So compensate for this the airline upgrades you to a first class and also notifies the hotel of your delayed arrival which has your digital identity stored in your profile. All of these transactions are obviously secured through Blockchain.
4. Consider a host of self-driving trucks(AI) who have to deliver goods to distribution centres using AI technology. A self driving truck is useless if it doesn’t charge automatically at a charging station. And for the charging station to charge the truck you need authentication, authorization and smart contract framework which works around this scenario. A ledger that can’t be tampered is employed to record the transaction in this smart contract which uses Ethereum (Blockchain). Once the goods reach the distribution centres they are then deployed to the customers through the drones (IoT) employed by the retailing company.
5. Fujitsu in July 2017 announced that they are developing an algorithm to assess heat stress of workers like security guards all using AI technologies and IoT. Fujitsu jointly developed together with Fujitsu Laboratories Ltd this algorithm in its digital platform that uses IoT to support on-site safety management. Previous algorithms could also detect heat stress levels using a sensor device to measure data like temperature and humidity. This new algorithm could assess heat stress levels that accumulate over time like employees who persistently work outdoors in the summer. This arrangement would be complete when the workers are insured and Blockchain is used to implement it. When the heat stress levels reach critical proportions then automatically the sensors on the body of the workers can detect it. Consider a worker’s health deteriorates due to it and the sensor had recorded a high temperature back then. Then the Blockchain smart contract would trigger the transfer of insurance money from the insurance company to the worker.
The success of the IoT and AI will not depend on the capability of the technology but rather on the security provided to data. It is nothing but a science exhibition where the prowess of AI and IoT devices is displayed without the security needs being met. Hence security is the main criteria to be fulfilled if these technologies can hope to really take off in a big way. Each of these technologies can leverage the strong points of each other and pave way for a vision where devices are securely transacting with each other making use of Blockchain technology and the data is processed through AI and Machine learning technology.
- Rising Demand for SMAC Skills in IT Firms
- Risk Management in Testing
- Qlikview Training – Business Intelligence Software