Many developing countries like Brazil, India, Bangladesh, UK have to deal with the electricity crisis. There could be many potential reasons behind this but one of the major ones is electricity theft. According to a research, UK loses more than 400 million pounds every year due to such malpractices. There are many people who try to save money by stealing electricity. This kind of practice is mainly adopted in rural areas or in some big factories or mills. But soon this problem could be solved by a team of software developers in Brazil.

The developers tested an AI algorithm on several household’s and the results were very promising. The developers are not only aiming at stopping electricity theft but they also want to get some crucial information like the peak electricity usage, the general trend in the electricity usage over the course of the year. All this information could be very useful for the local government and can help them tackle the problem of the electricity crisis.

The algorithm can recognize when energy use at a property was suspiciously low. Developers used the past 5 years electricity consumption data to train the AI. The algorithm will use the past data against the current consumption. This could help better target physical inspections of properties which could be a hectic process. According to the researchers, the current algorithm detects the malpractice with a 65% accuracy rate.

Now, researchers are aiming to implement this technology in commercial software that will be used in Latin America. If the algorithm could work as suggested by the researchers in real life then this could help developing countries boost their development rate. This would also mean that countries could use the extra money in another task like education, healthcare.

Image Source: New Scientist