Human-robot interaction and future industrial robotics applications

Abstract

Industrial robots, designed for performing operations quickly, repeatedly and accurately have a long heritage in the manufacturing industry, operating in relatively static environments and in large numbers. Trends in the oil and gas industry to improve safety and efficiency and reduce environmental impact suggest the use of industrial robotics. New developments in regions difficult or dangerous for humans to work in could be enabled with maintenance, inspection and repairs carried out by remotely-controlled industrial robots. This new application area highlights some difficulties with today's robots, as they do not adapt well to dynamic environments, do not offer rich human-robot interaction and are not simple for end-users to program. In addition, the oil and gas context presents a challenging work environment for robots as they are exposed to variable and often extreme weather and need to be safe for use alongside explosive hydrocarbons. As robots are introduced, issues of trust and accountability come to the fore as well as how they fit into organisational structures. If robots have too little autonomy, human operators will waste time attending to robots instead of attending to their work tasks. If robots are highly autonomous, situational awareness of plant activity is diminished. A balance needs to be struck to find a level of autonomy suitable for the task, the realistic capabilities of the automation, and the need to actively engage human operators in a constructive fashion. These issues also relate to what form the interface takes for remote or co-located robot control, as well as how information and activity is represented for remote operators. The contribution of this paper is a discussion of issues relating to human-robot interaction for future industrial robotics, in particular for the oil and gas industry

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 203 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.