Using AI to make the power supply more stable, Huijing Technology increases the income of customers by 20% Solar photovoltaics and wind power are renewable energy that Taiwan is currently actively promoting. In 2021, the solar power generation capacity will reach 7.92 billion. It is expected that the solar photovoltaic installation capacity will reach 20GW in 2025. In addition to the construction of large-scale workshops, the rooftop solar energy that is currently spread across factory buildings is also an important source of solar energy. In the face of scattered solar energy, AI technology can not only help solar industry real-time monitoring of the site, but also increase revenue. Due to climate, environment, placement angle, deterioration of module wiring... many reasons will affect the efficiency of solar power generation. The industry has to manage a huge project site and hundreds of solar panels at a time, which is not only burdensome but also inefficient. In addition, the cleaning of traditional solar panels is not only time-consuming, but may also delay the timing of power generation. Thingnario (thingnario) designs machine learning or Rule-Base models for different power generation abnormalities to find out the reasons for the decline in power generation and prevent them early. Chen Weiting, product director of Huijing Technology, said that the company mainly focuses on the vertical fields of the energy industry, especially distributed renewable energy, including photovoltaics, wind power, and energy storage. At present, in the solar energy industry, the installed capacity of 1G has been monitored, and there are more than 2,000 distributed case sites, which effectively help the case field to increase revenue by 12-20%. Because the data of the solar field will be continuously transmitted back into the system, for general monitoring software, the vertical and horizontal expansion of the database is a big problem, but Chen Weiting said that because the team has previous project experience in video surveillance products , so it still has an advantage over other competitors in terms of technical issues of time-sequential massive data processing. The time for the energy transition has arrived As for why you choose to invest in the development of solar maintenance and monitoring systems? Chen Weiting said that at that time, the team mainly focused on industries that had transformation needs and were growing, because only in this way could they have the opportunity to enter the market, and the energy industry was one of them. In terms of power distribution, with the rise of renewable energy, the centralized system of large power plants in the past is slowly turning to decentralized management. "The future trend of the world is to move towards microgrids," Chen Weiting believes that this decentralized system Energy management will require IOT technology, and the field of industrial control and the processing of large amounts of data are exactly the areas that the team is good at. Huijing Technology was established in 2017 and invested in the development of solar maintenance and monitoring systems, and then entered the overseas market in 2019. Chen Weiting is also very pleased that this system has started to make profits within three years of its launch. She said that this product is only the first step, and the next step they will challenge is to help establish a stable grid service. Taking Taiwan's electricity as an example, it is not very accurate to say whether the electricity is enough or not. More importantly, it is stable and unstable. Chen Weiting believes that when the proportion of renewable energy increases to 20% in the future, it is even more necessary to grasp power changes in real time and deploy power in a timely manner. The current practice of Taipower is to prepare as much electricity as the user needs. But how to judge how much electricity the user needs? Chen Weiting said that at present, the data of the past few years and weather factors are mainly used to estimate, but because there are too many variables in the middle, real-time collection and correction of power consumption through IOT can make power deployment more efficient. All can be optimized.
Forum Role: Participant
Topics Started: 0
Replies Created: 0