In the context of big data， strategies to improve the self-directed learning ability of engineering college students in local universities were studied from four aspects： universities， teachers， teaching methods， and assessment and supervision mechanisms. The results indicate that in order to cultivate the self-learning ability of engineering college students， firstly， universities should formulate favorable policies and create a good self-learning environment； secondly， it is necessary to leverage the subjective initiative of teachers based on big data technology to correctly channel the learning motivation of college students， help them determine learning goals， design learning plans， and enhance their interest and sense of responsibility in study； thirdly， we must adopt multidimensional teaching methods to improve the classroom learning efficiency of engineering college students， stimulate their enthusiasm， initiative， and creativity， and enhance their sense of fulfillment and confidence； fourthly， a scientific assessment and supervision mechanism should be established to correct the learning attitude of engineering college students. Ultimately， we can improve the teaching results of engineering college students and cultivate high-quality engineering and technical talents.