AI and Data Acquisition - A Case Study
Artificial intelligence offers many opportunities for data acquisition systems. It can help reduce costs by removing the need for physical sensors. AI systems can be easily scaled up to handle large volumes of data from many sources. Its machine learning leads to continuous improvements in performance, making data acquisition systems more adaptive and responsive to changing conditions. AI allows organisations to easily customise data collection systems to meet their specific needs, whether it's in agriculture, transport or other industries.
One of example of how artificial intelligence is advancing data acquisition is in computer vision: analysing and understanding visual information and translating it into quantifiable data. For instance, in the new systems now being developed to count the number of passengers on public transport. We'll explore this a little further to show how AI can improve traditional data collection systems.
Traditional Methods vs. AI-Driven Systems
Traditionally, public transport systems have relied on manual ticketing, periodic surveys or basic sensors to gather passenger data. These methods, while effective to a degree, have their drawbacks. Manual methods are prone to human error, while basic sensors can't differentiate between a person, a piece of luggage, or even the vehicle's own movements.
This is where AI steps in. With its advanced image and motion recognition capabilities, AI can be trained to recognise and count passengers under various conditions, offering a significantly higher accuracy rate.
Detecting Door Movements
One of the biggest challenges in counting passengers is determining when to start and stop the count. By leveraging AI to detect when doors are open, we can effectively begin counting only when passengers are boarding or alighting. This eliminates potential miscounts that could occur when the vehicle is busy and doors are closed, ensuring that only relevant data is captured.
Vehicle-Specific Training
Every vehicle type - be it a bus, train, tram, or ferry - has its unique set of challenges. Different lighting conditions, types of doors, and passenger behaviours can all impact counting accuracy. Fortunately, with AI, systems can be trained specifically for each vehicle type.
For instance, ferries may have larger doorways and more open spaces, while trams might have frequent stops with rapid passenger movement. AI can be tailored to understand these nuances, ensuring precise counts irrespective of the transport mode.
Adapting to Environments
From the dimly lit interiors of a nighttime bus to the sun-drenched decks of a ferry, the lighting conditions can vary drastically in public transport. With AI's ability to be trained using vast datasets, it can adapt to these changes, ensuring that passenger counts are not compromised due to environmental factors.
Additionally, certain cultural or regional behaviours - like passengers clustering near doors or moving in groups - can be incorporated into the AI's learning, making the system globally adaptable.
Future of Data Acquisition
As we move towards smarter cities and systems, AI-powered solutions will undoubtedly play an integral role in shaping the data acquisition landscape.
by David Collins, Retail Sensing. For more information contact sales@windmill.co.uk.
