Thanks to popular media, when people hear the words Artificial Intelligence (AI) they think of sci-fi movies like 2001: A Space Odyssey, The Terminator, or even The Matrix. They’re usually led to believe that AI will result in global catastrophe (dystopia) or a future where all our problems will be solved (utopia). The truth, however, is somewhere in between.

A Brief History of AI

AI isn’t new, it’s been around since the 1950s. The Stanford Workshop of 1956, which is regarded as the birthplace of AI, set off a golden age of AI marking significant advances in the ability of computers to solve problems like algebra, geometry and speaking English. The initial hype was unmet - despite significant funding and progress, AI was limited by available, cost-effective compute power and data. This “AI Winter” ended in the 1980s as programs such as “expert systems” became popular with corporations leading to a brief boom in AI. By the late 1980s, again due to unachievable expectations and cost of these systems, the AI field deflated again. From rebased expectations AI has been steadily, and slowly, making progress moving into the modern era.

What is AI?

AI is simply a computer program that can accomplish difficult tasks, tasks that humans would say require intelligence. Many fields come under the umbrella of AI, the most important of which today is Machine Learning (ML).

What is Machine Learning?

ML and AI are often used interchangeably and this can be confusing. ML is simply a computer program that is programmed to learn - as you give the program more and more data it changes itself or “learns” and performs better.  ML thrives on large quantities of diverse data, generally the more the better. Often data that comes from machines is already structured (e.g. numbers in rows and columns in a database) making it easy to use. But when it isn’t (e.g. a photograph), it must be labelled or organized in a structured manner - a task usually performed by humans.

What isn’t Machine Learning?

A software program is simply a set of instructions to turn a given set of inputs, into a set of outputs. Traditional software programs are hard-coded sets of instructions - once the program runs, the instructions don’t change. ML is a software program which can change itself as it uses data - in other words, it learns and improves over time.

The new era of AI

Fast forward to today, AI has generated a tremendous amount of publicity and attention. The three key drivers of the modern era of AI are:

• More data is available as IoT devices are increasingly commonplace
• Affordable and scalable in the cloud Graphical Processing Units (GPUs) now make compute power cheaper and more efficient for AI programs
• There are now widely available knowledge, tools and frameworks available for developers. Tasks that required a team of researchers 20 years ago can now be accomplished by an individual in a few days at substantially lower cost.

The new AI age is achieving massive advances in nearly every major industry including manufacturing, automotive and advertising and it is here to stay. At Flow Labs, we’re now bringing it to the Traffic Management industry.

So what does this mean for you?

Terabytes and terabytes of data are collected from your transportation network on a daily basis. This data is collected from roadside detection, your traffic signals and even the vehicles traveling on your roads. Much of this data is probably sitting in your servers, or a data warehouse right now, waiting to be used. This data provides the fuel for AI to help you and your teams make valuable, impactful decisions from Traffic Signal Control to Transit Planning to Traffic Incident Management creating the intelligent transportation system.

As an organization you should be asking yourselves:

• Do we have a problem? Is there a problem without a solution currently? Is something taking us a lot of labor time or cost?
• Do we have data about the problem? Where it happens, when it happens, what happens. Is there data we have but don’t know how to use? Is there more data that we can handle?
• Do you have enough data about the problem? Different problems require different amounts of data, just ask.

Flow Labs technologies like Predictive Traffic Control (PTC) enable you to spend less, deploy faster and see results quicker. Contact us to schedule a demo today.