Artificial intelligence for traffic planning and safer surroundings
Flowity is the name of a new product developed by AFRY. By analyzing video, Flowity can map how vehicles, people, or other entities move in different environments.
There is a lot to be gained by analyzing data on how vehicles and people move. For one thing, traffic can be planned smarter, both when creating the infrastructure, but also during roadblocks or accidents. Also, during a large event in a city, the data can contribute in terms of safety – making sure that not too many people get stuck in the same place at the same time.
Flowity uses a technique that is based on artificial intelligence, and works as a platform that analyzes video to track how objects move. The data is then used for creating smoother flows, for instance in traffic or in a store. From a sustainability perspective it can also be used to achieve smarter energy management and efficiency through smarter control of air conditioning, lighting and electrical distribution.
It can be run in real time, where the collected information is processed instantly, and then immediately discarded. As such, Flowity is well suited to be used in settings where it is crucial to collect fully anonymized data to conform with data regulations, such as GDPR.
”What makes Flowity unique is its use of deep neural networks to detect and track objects. Flowity is able to recognize the actual object, a feat that has been accomplished through enabling the algorithm see and train on a substantial collection of images which contains the object in question. This results in a versatile product, that can be adapted to recognize any type of object and that is more resilient to noise, such as sudden changes in brightness” says Roger Tedblad, software developer at AFRY.
An example on how Flowity can be used
Imagine a store owner that wants to increase the sales. With Flowity, he can identify different characteristics of his buyers, be it by their age, gender or motives, and track the movement patterns of the clientele, determining whether they are in a hurry or not. Are they dawdling shoppers or are they looking for a specific good?
From the data, the store owner gets an indication on how to plan the store, at what hours specific goods should have special positions and keep track on failed purchases due to understocked shelves.
AI technology for real estate
Currently, AFRY carries out a project for a major real estate company, where Flowity is used to count the numbers of visitors in the retail business. AI is also used for mapping energy use and creating self-learning functions to optimize the energy usage in the customer’s properties.