Fig. 5. SmartSite platform influencing the entire supply chain in asphalt road construction. 

develop new methods to integrate quality-related data into the ma- chine control, specifically in this article for the asphalt compactors.

The construction process control is supplied directly with real-time data from the construction site. The monitoring of the process execution as well as the visualization is achieved via smart phones and smart tablets — the future working devices of the machine operators and of the site manager.

Fig. 6 shows the different aspects of the construction process monitoring and control developed and used in SmartSite. While the state-of-the-art desktop BIM applications will still be available and used – especially in the pre-construction planning phase – the monitor- ing during the construction phase will be shifted to mobile devices, car- ried out by the human actors on site. The construction machines on site will be tracked via high precision GPS. The positioning system used in SmartSite is TOPCON's proprietary mmGPS P63. The system uses GNSS with a position accuracy of 10–20 mm and a PZL-1 zone laser transmit- ter for a height accuracy of 2–3 mm. Each inpidual compactor has a PZS-MC laser receiver and a GX-60 control box installed.

Further sensory data will be collected by both the machines and additional sensor clusters like a weather station. The sensors on the in- pidual machines allow for the collection of various data like current asphalt temperature and stiffness as well as data on the operation of the machines themselves. Interfaces linking off- or onsite office data with the outdoor construction field allow real-time data collection and control via a central access point. Performing some semantic

enrichment like mapping the asphalt layer related data to a digital map of the construction site becomes possible. Schedule and cost prog- ress reports and automatic billing becomes feasible, among many other benefits to several stakeholders involved in the asphalt road construc- tion process. Table 2 displays some of the main innovations introduced by SmartSite research project and how they pertain to autonomy.

The following specific equipment, machine, and plant innovations will  be made through  a successful  SmartSite project:

• Plants: Current asphalt plants are not sufficiently integrated into road construction information systems. The delivery schedule is typically predetermined with preset total volumes per day. As such the batch plant becomes a bottleneck in the optimization of the construction site as a whole since deliveries cannot be adjusted based on as-built information. Furthermore, data transfer often still relies on outdated methods such as E-mail and fax for orders and physical bills of deliv- ery. SmartSite features fully integrated batch plants. This not only drastically reduces the time spent at the plant for each truck but also allows for dynamic adjustment of delivery times. By utilizing live ve- hicle tracking the pending arrival of a truck can be transmitted to the batch plant well in advance.

本文介绍了SmartSite研究项目的概况，该项目采用机器学习，决策理论和分布式人工智能来设计和测试用于沥青道路施工的多代理系统（MAS）。SmartSite重点强调传感和通信技术，将实时自动化信息交换集成在道路建设的供应链中。作为更大的SmartSite项目的一部分，本文介绍了压实机的新型实时路径规划系统，并介绍了在复杂的模拟和恶劣的施工环境中进行评估系统的几个模拟和现场现实实验的结果。该系统基于信任-期望-意图（BDI）软件代理和实时感觉输入来操作。新开发的集成和信息丰富的过程有利于沥青压实操作员，因为他们现在能够控制他们的机械，并对变化的环境，材料相关和过程相关的干扰或变化做出反应。这提高了沥青材料的输送和铺设的质量，防止压实机过度压实某些路段，在操作生命周期阶段期间增加道路的铺路寿命;重新调整现场经理的工作任务，减少施工预算和时间表。该系统在实际操作中操纵沥青压路机的能力也使其成为全自动沥青压实机的重要一步。