FW-HTF-R:Collaborative Research: Partnering Workers with Interactive Robot Assistants to Usher Transformation in Future Construction Work

建筑业是一个价值10万亿美元的行业，在全球雇佣了大约1.8亿工人。然而，建筑工作的未来正处于十字路口。首先，相对于其他行业（如制造业），建筑工作的生产率一直停滞不前，而且该行业在采用影响效率的创新方面一直很慢。其次，很难用更年轻、更多样化的劳动力来抵消老龄化和退休的劳动力，导致劳动力供应无法满足不断增长的需求。这主要是因为建筑工作往往是体力劳动，导致职业危害，往往迫使工人提前退休。机器人化被认为是解决这些问题的一个潜在方法。然而，建筑工作的非结构化性质带来了一些技术、社会和经济障碍，阻碍了建筑行业直接采用和整合这些创新。对于建筑工人来说，只有当机器人技术能够让他们在工作中释放激情，同时避免与体力需求相关的慢性疼痛和健康问题时，机器人技术才能带来变革。该项目调查了建筑工作是否可以被视为人类与机器人的合作伙伴关系，人类工人在规划工作、培训和监督机器人助手以适应所呈现的工作空间条件并执行有用的工作方面发挥关键作用。项目团队正在整合交互式任务学习、混合现实和强化学习方面的先进技术，使建筑工人能够通过直接的物理交互和虚拟监督和培训，与机器人助手自然协作。为了让这种人机共生的伙伴关系惠及建筑工人，并使其得到广泛部署，工人们需要掌握新的技能。项目团队正在探索新的教育和专业发展计划，以支持工人提高技能和终身学习的愿望，并为不同能力的人开辟成为建筑劳动力中富有成效的成员的途径。与行业合作者的紧密合作将为项目活动提供信息，并为测试和评估提供进入建筑工地和培训设施的机会。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Construction is a $10 trillion industry that employs about 180 million workers worldwide. However, the future of construction work is at crossroads. First, productivity in construction work has been stagnant relative to other industries (e.g., manufacturing), and the industry has historically been slow to adopt innovations that affect efficiency. Second, it has been difficult to offset the aging and retiring workforce with younger and more diverse workers, causing the workforce supply to fall short of rising demand. This is mainly because construction work tends to be physically strenuous leading to occupational hazards that often force workers to retire early. Robotization has been suggested as a potential solution to these problems. However, the unstructured nature of construction work presents several technical, social and economic impediments that hinder the direct adoption and integration of such innovations by the construction industry. For construction workers, robotic technology can only be transformative if it allows them to channel their passion for the work while avoiding the chronic pain and health outcomes associated with its physical demands.This project investigates if construction work can be conceived as a human-robot partnership, where human workers play the critical role of planning the work, and training and supervising robotic assistants to adapt to presented workspace conditions and perform useful work. The project team is integrating advances in interactive task learning, mixed reality, and reinforcement learning to enable construction workers to naturally collaborate with robot assistants through direct physical interaction and virtual supervision and training. For such a symbiotic human-robot partnership to benefit construction workers and result in widespread deployment, workers need to be equipped with new skills. The project team is exploring new educational and professional development programs to support worker aspirations for upskilling and lifelong learning, and to open avenues for people of diverse abilities to be productive members of the construction workforce. Tight-knit partnerships with industry collaborators will inform the project activities and provide access to construction work sites and training facilities for testing and evaluation.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.