Artificial intelligence has been widely used in art education and learning due to its quick progress. Any creation made with the help of artificial intelligence is referred to as art design. It covers works generated independently by AI systems and works created in collaboration with humans and AI systems. The objective of directing the invention of environmental art design thinking is to stimulate students' learning and innovation abilities and teach students how to put design ideas into effect. Despite the progress of smart technologies, there are several challenges in increasing the teaching capabilities of technical art design courses, such as the influence of different variables, the absence of quantitative research, and the imperfection in the index system. In this paper, the Artificial intelligence-based Art design and teaching (AI-ADT) method in colleges increases the capacity to adapt to AI-oriented art education, establish intelligent teaching methods, and improve AI-oriented art teaching art knowledge and environments. The widespread application of artificial intelligence in design education has become a trend in development. Self-Learning Systems are software that incorporates machine learning techniques to allow computers to learn from and make judgments based on data without the need for explicit programming instructions.The art design profession should confirm and actively adapt to this development trend. Modify the original teaching mode, invent their teaching methods, continually enrich the teaching methods, enhance the quality of teaching, and constantly foster high-quality art design talents in the new age. AI-ADT investigates the optimization of the art design curriculum system in higher education institutions in the context of artificial intelligence. The experimental results show that the proposed method develops smart teaching (98.1%), flexibility (96.5%), performance (93.6%), participation (94.9%), interaction (95.1%).

Industry 4.0 is derived from the concept of German industry, and its essence is to use technology for efficient production. However, the evolution in wireless technologies has opened the door to a new class of plant automation architecture that offers adopters a significant strategic advantage. Under the historical background of advocating low-carbon production, energy conservation and environmental protection, and sustainable development in various fields, the awareness of ecological energy conservation and environmental protection has gradually become a mainstream ideology. This paper principally studies the application of energy conservation and environmental protection green decoration materials in public architectural engineering projects to show the advantages of energy conservation and environmental protection green decoration materials in architectural decoration construction and analyzes the future development direction of public architectural decoration in China. Therefore, it is of great significance to improve the public environment and develop new ways of housing renovation with sustainable development goals to create a sustainable Industry 4.0 combining environmental protection and sustainability.

Outdoor positioning can often achieve accurate positioning according to GPS and mobile phone signaling, while indoor positioning is difficult to meet the needs of practical application due to the limitations of satellite reception. In order to effectively solve the problem of large error in the individual positioning strategy in the indoor environment, this paper applies multisensor in the multisource information fusion indoor positioning system. By using the positioning results of multiple sensors to limit the range of geomagnetic matching for combined matching, the matching error can be effectively reduced. Then, the global optimal value of indoor network is calculated by using the multi-information data fusion algorithm, which can optimize the initial value and threshold of the multi-information data fusion algorithm, improve the network accuracy as much as possible, and accelerate the convergence speed at the same time. After completing the optimization processing, the indoor network can obtain the combined positioning and predicted positioning results, so as to facilitate the fusion training to the actual position coordinates, and finally obtain the optimal positioning results. The simulation results show that the mean square error predicted by the multi-information data fusion algorithm calculated by the multi-information data fusion algorithm can be effectively reduced by 76%, and the fusion positioning accuracy can be improved by 48% compared with the accuracy of a single positioning strategy. The method proposed in this paper effectively improves the positioning accuracy, indicating that the positioning performance is better.