The information provided on this page is for reference only. Please refer to the latest information published by the partner company.
Predicx aims to applying advanced predictive algorithms to improve data perception and data analysis capabilities. Our enhanced predictive algorithms and models can be applied to multiple scenarios to solve the problems we face, such as indoor air quality monitoring: data acquired by low-cost IoT sensors, to simulate the functions that can only be realiszed by a special and complex environmental monitor, and can continuously monitor the impact of continuous simulated fungi, bacteria, air pollutants, etc. on the target environment.
Our AI modelling is suitable for processing time-series streaming data, and uses deep learning models to predict targets, such as energy consumption prediction of industrial and commercial districts. This technology has been approached and applied by many customers.
Predicx's product solutions combine a number of advanced technologies to enhance the data perception, data analysis and model prediction capabilities of IoT business applications.
For example, our indoor air quality monitoring solution uses low-cost IoT sensors to collect indoor air quality data. These IoT sensors use standard transmission technologies such as Zigbee and LoRaWan to transmit data streams to the local edge computing for preliminary storage, calculation, and monitoring, meanwhile the data can be further transmitted to the cloud platform for storage, analysis, and continuous model training.
Our energy consumption prediction technology can be applied to energy management in various scenarios, such as identifying the energy consumption pattern for industrial and commercial buildings, then predicting the energy demand, scheduling the charging and discharging of local battery storage system, and optimising response strategies for peak loading shifting, etc., it can also be applied smart management for the off-grid new energy system, improving the reliability of energy supply, and prolonging the service life of energy storage batteries.
Our prediction technology is optimised based on mainstream machine learning algorithms such as XGBoost/LightGBM, RNN/LSTM, which can meet the prediction capability requirements of different business applications. With the continuous expansion of future applications, we will further modularise prediction technology and packaging API Interface call, which will be more convenient for customers to adopt our predicting technology to realise business value.