AI has always been shorthand for whatever has not been accomplished yet. But what distinguishes AI from what may be described as just automation?
Regardless of how it is built, AI is differentiated from automation by its surprising outcomes. Those surprises come because the software does more on its own, and because it pursues goals rather than slavishly follows pre-defined rules. We propose a simple definition: AI is software with narrow but humanlike intelligence and autonomy— for example, planning road transport without the use of hand-crafted instructions.
So when we talk about AI for trucking we mean a system that comes out of the box with the know-how, flexibility, and tactics of the most skilled transport planners. AI is software that gets to the result without being told how to.
AI also implies more autonomy over decision-making. We use a scale, shown here, to explain the gradations between manual and autonomous dispatching. AI for Trucking means software that achieving goals without being told how to do so, and with a high autonomy of decision making.
|Grade||Software Autonomy in the Decision|
|0||No system involvement.|
|1||System collects data inputs,|
|2||which are exhaustive, then|
|3||calculates their implications, and|
|4||gives a list of options,|
|5||which are exhaustive, or|
|6||narrows down the options to a few, and|
|7||suggests a decision that the user could override, or else|
|8||executes the decision automatically after a period of time, or|
|9||executes immediately, then notifies the user, or|
|10||the system decides everything, informing the user only for exceptions.|