Abstract:

In recent years, Japan's third AI boom has begun to settle down, with practical initiatives taking root in industry. However, for small and medium-sized enterprises (SMEs), not large corporations, the barriers to introducing AI technology into their core business remain high. The reality is that they lack the financial and time resources to train AI specialists in-house. Particularly in industries outside the ICT sector, understanding of AI itself remains limited. While mastering applied technologies requires learning fundamental theories, as previously noted, finding the time for this is often difficult. Introducing AI into operations heavily reliant on manual labor or industries that depend on veteran expertise is especially urgent.

This paper describes a training program for Japan's construction industry that includes lectures on AI and a comprehensive process leading to the practical implementation of AI technologies. The program targets medium-sized private construction companies and technical associations whose members are construction engineers. Both types of organizations possess limited specialized knowledge regarding ICT and AI, and the program began with the phase of considering what AI could achieve within their own companies. Furthermore, the various materials and practical tools used in this training primarily draw from those that have been well-received in lectures and experiments the author has conducted at their affiliated university and other institutions. This practice also aims to verify whether these tools prove useful for different target audiences: students and corporate professionals.

In practice, we presented attendees with classification methods using free machine learning services and conducted hands-on exercises where they classified images, audio, and other data within their own skill levels. This service enables nearly full mouse-based operation, allowing users to easily verify AI results without writing code. Through this practice, attendees assessed their own skill levels and achieved satisfactory results within their capabilities. We will now examine the outcomes and effects of this approach.

Keywords:

Training Programs for Construction Industry, Artificial Intelligence, Image Processing, Classification, Google Teachable Machine.