Electronic products are often subject to various harmful effects from the surrounding environment during storage, transportation, and use, such as affecting the working performance, reliability, and lifespan of electronic products. The environmental factors that affect electronic products include temperature, humidity, atmospheric pressure, solar radiation, rain, wind, ice and snow, dust and sand, salt mist, corrosive gases, mold, insects and other harmful animals, vibration, impact, earthquake, collision, centrifugal acceleration, sound vibration, sway, electromagnetic interference, and lightning.

The research on environmental factors mainly solves two basic problems: ① how to obtain objective data on these environmental factors; ② How to handle this data. Data on objective environmental factors can usually be partially obtained from meteorological and environmental protection departments, but more must be obtained through actual measurements. In order for the measured data to have both reliability and typicality, in addition to a comprehensive investigation and testing plan, there must also be instruments that can continuously, quickly, and record multiple points. If the objective environmental data obtained has a sufficiently long recording time, statistical analysis can be conducted based on the frequency of occurrence. For products that require special reliability, extreme values of objective environmental data, or even statistical inference, can be taken to ensure that the product is foolproof in use. For products that require high reliability, objective environmental data with a probability of 1% can be used. For products with general requirements, an objective environment probability of 5% or even 10% can be considered. If the recording time of objective environmental data is not long enough, mathematical and statistical knowledge should be used to process it. For example, microclimate survey data can be extended using correlation methods to infer possible historical data; For example, the actual measurement and investigation data of mechanical vibration can be calculated using envelope method, power spectrum analysis method, or time series modeling method to determine the probability of various probability values, and then the required data can be obtained based on the reliability requirements of the product.