Internet of Things (IoT) is a key technology for smart city, industry, and agriculture and anything that can benefit from improved understanding of the state of the world around us. Logistical decisions to help determine where to expend effort and resources can be greatly optimised when knowledge about all the moving and non-moving parts of the system are understood.
The game changer is in the form of data-driven intelligent decision making, enabled by IoT. Intelligence helps people make smarter decisions that are better for the prosperity of people and the planet.
Reduction in the consumption of precious commodities such as fuel, fertilizer and water, not to mention the reduction is waste and pollutants are game-changing for the lives of humans, plants and animals.
What would you do if you had new super powers to know when something would happen in advance of it occurring? What if you could remotely manipulate the environment from within your body to outer space?
This is the power of Internet of Things.
Intelligence is evident when the actions of an agent demonstrate thoughtful consideration of not just available information, but the context that information was derived within, as well as the pros and cons of acting on one of multiple options. Living things show various levels of intelligence, but it is those that are able to rapidly learn and adjust to new information that demonstrate superior intelligence.
Internet of Things can offer intelligent options for action by humans or machines by providing improved information about the world and a way to analyse it. The information is required to evaluate an event, its components (things and people), location and time, and the situation that is known to cause the event. This can be analysed by computers along with the events preceding the event to signal patterns in the sequence and type of events can indicate future consequential events and can be surfaced for consideration.
IoT provides greater knowledge about the present situation which enables a greater ability to infer, with a greater degree of certainty, what will happen next.
Machines are perfectly equipped to demonstrate intelligent behaviour when the rules are known and data is available about the present situation, the preceding events and the location of those events.
For many problematic situations there are tried and tested responses. For example, if motor temperature is unusually hot and audio emission is unusually noisy, then turn it off to avoid a fire. These problem/solution pairs are good in deterministic systems. They are great when based on physical closed systems that are controllable and shielded from external events.
For problems that don't fit into the closed system paradigm, generally involving large and rapidly changing or undetermined variables, a predetermined if-this-then-that decision cannot be feasibly applied to solve the problem. For these "hard" problems, computers can analyse options to determine the best choice to inform a human who can then determine the risk of taking action, or wait while more data is gathered and analysed.
Not all IoT intelligence is useful and humans in the loop will be critical for hard problems.
Analysis of options can be done through ultra-fast computer simulations. Alternatively, without a human-in-the-loop the computer could cautiously and safely test several options to learn through feedback which option is viable before committing to actioning the best option. In these circumstances where machine learning is leveraged, it is less important to know beforehand how to solve a problem, rather it is important to set the safe boundaries within the solution must operate.
With Internet of Things, we can collect data to make a smart decisions because a greater number of alternatives have been considered.
Humans cannot consider millions of data points and simulate outcomes of millions of alternative actions in minutes, but humans can tasks machines to do this to augment their intelligence.
Today’s machine data analytic technologies are perfectly equipped to automatically and intelligently support your decisions. Machine Learning Artificial Intelligence considers the present situation, location and time-based context in relation to preceding events.
Today, we fuel up heavy garbage trucks to drive up streets where the recycling bins might be mostly empty or not put out at all. With IoT the bins can communicate with the street and the truck to save unnecessary trips.
Furthermore, the data collection about how much garbage each bin has, the contents, the smells, temperature, or images can be analysed for opportunities to improve benefits not just for the local garbage collection, but for the environment and society.
Today, our supermarkets dispose of spoilt food because they don’t know that we had pantries and fridges full of that food already.
With the ability to acquire and interpret information about the events in the environment around us, we can cross-analyse this data along with related data such as weather, satellite and other available external data sets, using computers to inform our decision making.
We can predict events and intervene before they happen. For example, don't water if it's about to rain. Don't leave the office yet because the route home is congested. Open your ice cream store at 6 instead of 7 on Tuesdays and increase your revenue by 20%. Create a Pokémon Go poke-stop in your shop and increase your revenue by an 40%. Hold 10% extra stock to increase your customer satisfaction by 15%.
There are many business models that can take great advantage of IoT solutions to realise the benefits.