Artificial intelligence (AI) puts computer systems to tasks that once required human intelligence. AI and automation will augment, improve and change the way of doing work, thereby fundamentally reshaping the workforce. Many customers experience AI in their day to day lives through their virtual assistants like Siri, Cortana and Alexa. A few cases to exemplify its diverse usage –
- Robotics uses AI
- Autonomous vehicles depend on AI to sort the information their sensors gather
- Online chatbot services use AI to greet people and resolve their issues
Two common terms frequenting around AI topics are –
Supervised Learning – Which is the most common way people train AI systems using data and guiding the system by understanding the distinctions. e.g Providing pictures that show animals and pictures that don’t to learn the distinctions and apply the learning to a new picture.
and Unsupervised Learning – Where systems start with data that mean nothing to them and identify patterns on their own.
Deep neural networks (DNNs) provide architecture to an AI system. These networks have multiple layers – the more complex a problem, the more layers. DNNs have an input layer, an output layer and hidden layers in between where the difficult work gets done. Nodes in one layer connect to nodes in others. Each connection is weighted, which creates both weak and strong links. Weaker links produce undesired answers during training and don’t pass along fruitful information. As developers train networks, the weights adjust to reach an optimal level.
While most of the capabilities of AI are still under research and far from being commercially utilized, we can categorize those capabilities under:
AI capabilities for information capturing – this includes speech recognition, image recognition, search and clustering. When machines capture information, they convert unstructured data (big data) to structured data. This requires speedy processors and a lot of training.
AI capabilities to understand what is happening – this includes Cognitive reasoning, optimization, prediction and machine understanding. It includes a degree of understanding which is supposed to be AI’s ultimate capability, and it requires machine’s ability to be consciously aware of what it does or thinks.
AI and Internet of Things –
AI also comes into play in the Internet of Things (IoT) when devices transmit data directly to each other. When billions of devices transmit data, this generates massive big data – a natural place to implement AI. When AI can’t complete a task, humans intervene, such as in crowdsourcing or in cases when a task exceeds a system’s capability – say, reading handwritten text etc.
IoT and their applications across industries –
The concept of Internet of things (IoT) usually includes both technology and services that is based upon connected devices and the usage of the collected data. IoT pushes towards networked business models from single firm business models. Following are a few cases that are early movers and have envisioned IoT usage and business models that have a potential to bring true digital disruption across industries –
The Connected Trashcans – Connecting trashcans to a monitoring system has become a solution (trashcans raise an alarm when they are full) to avoid emptying half-empty trash cans, thereby saving time and the environment. One provider of such a solar panel-based trashcan solution is Big-Belly. These items are termed as Urban Furniture and since they are a connected system they can be augmented with features like – offering people WiFi services, providing information to visitors about the local area etc. Other examples of Urban Furniture include: lamp posts, bus stops, ad signs, and billboards.
The Connected Bolt – A Swedish start-up company,StrainLabs, has developed a connected bolt which has a sensor embedded into a cavity in the head of the bolt which monitors load and temperature. When it detects that a bolt is about to come loose, it alerts the user; so that, preemptive action can be undertaken. Inspection today is a manual activity performed in the field, posing several challenges with platforms that have a hundred thousand bolts in hard-to-reach places that need to be inspected periodically. With the advent of the connected bolts StrainLabs has ventured into selling a service (the service to monitor the screw connection) instead of products, opening up a new business in the area of – monitoring and inspection.
Connected Vehicles – Manufacturers of cars and transport vehicles have looked for years into the possibilities with connected cars, self-driving (or driver-assisted) cars and connected vehicles. Companies such as MAN, Volvo, and Scania have developed and have used solutions where one can track different types of data: for example vehicle location, engine, driving, and driver information. Scania has been possibly moving towards “transport as a service” instead of selling vehicles – with a large fleet of connected trucks that are delivered to different trans-port companies enabling common interfaces and third parties to access vehicle data.
Digital and Smart Locks – Digital locks or smart locks are part of a growing market, mainly focusing upon offices and homes as an access system to real estate buildings, and in various applications within other industries. For the individual private house owner, the digital lock is a fully adequate solution that brings keyless convenience to unlock the door by choosing from a range of digital door locking technology like – fingerprint, smart card, mobile phones or keypad – also paired with alarms to offer additional security from burglary, vandalism and fire.
The Connected Service Box or Delivery Room – The company Qlocx offers solutions where logistics companies and users have access to common delivery “boxes” or rooms. Qlocx develops these smart delivery containers and spaces to ensure that one receives goods or deliveries without physically needing to be there. Qlocx has also developed smart digital locks – One can open the device with a mobile phone and hand over the digital keys to anyone. Hence, one can share even car keys with their neighbors, by giving out a digital one-time key.
Above, we have tried to identify the significance of Artificial Intelligence and many cases using IoT (augmented by AI capabilities) in different industries: industrial IoT, smart energy, smart homes, smart cities, transportation, healthcare and social care, sports and well-being.
However, it is still into the nascent stage with a number of business opportunities and obstacles standing before their commercial scale adoption.