Laser Scanner based Probabilistic Predictive and Adaptive Safety Control Algorithm for Construction Vehicles

Abstract

Over the last decade, fatal accidents in construction sites due to operators carelessness and blind spots around construction vehicles have been accounted for 34.9 percent of all accidents. And the fatal accidents such as collision and sticking have been accounted for 72.2 percent of accidents due to operators carelessness and blind spots. For this reason, various safety systems such as blind spot detection and safety alert systems have been developed and marketed by construction machinery makers. Furthermore, there are various ongoing projects which are trying to achieve the zero fatality in construction sites. Several research teams around the world are continuously advancing the field of safety system. And some of major construction machinery makers have been researching to integrate individual safety system for the enhancement of safety. Volvo is trying to produce high quality, safe and environmentally-friendly machines that meet customer needs and enhance product offering in todays highly competitive global marketplace. And Volvo is developing both passive and active safety solutions that include everything from ensuring safe machine entry and exit to intelligent machines. In the future, Volvo expects that machines will be able to detect obstacles and humans and therefore instinctively avoid collisions. Caterpillar has undertaken researches to develop Safety system for zero-incident performance technology for achieving operational excellence through safety culture development. From a careful review of considerable amount of literature, active safety technology has the potential to reduce the fatal accidents such as collision and sticking accidents, and increase the safety of working in construction sites. However, the current state-of-the-art in safety technology for construction equipment requires precise, expensive sensors such as differential global positioning systems, and highly accurate inertial navigation systems and scanning laser rangefinders. While the cost of these sensors is going down, integrating them into safety system for construction vehicles will increase the price and represent yet another barrier to adoption. Therefore, this dissertation focused on developing a fully active safety control algorithm which is capable of collision avoidance in various scenarios. Chosen sensor configuration is closer to current automotive serial production in terms of cost and technical maturity than in many autonomous vehicles presented earlier. Mainly three research issues are considered: a predictive environment monitoring, an adaptive decision, and an excavator control. In the remainder of this thesis, an overview of the overall architecture of the proposed safety control algorithm for construction vehicles is proposed and the simulation-based performance evaluation which shown the effectiveness of the proposed safety control algorithm. The effectiveness of the proposed safety control algorithm is evaluated via computer simulations. The results show the good performance with on working scenario with static and moving objects.