Lead Institution: University of Sheffield
Industry Partners: ELG Utica Alloys, Advanced Manufacturing Research Centre
Project Team:  Prof Rob Gaizauskas (PI), Dr Emma Barker, Dr James Law and Dr Sam Fernando (all Sheffield)
Project Duration:  12 months (01 April 2017 – 31 March 2018)

Information Sheet

Research Challenge
Across the industrialised world there is broad agreement that we are on the brink of revolutionary changes in manufacturing, brought about by the convergence of industrial production and information and communication technologies.  Amongst the many emerging capabilities envisaged to form part of this revolution are: (1) human-robot
co-working, where the complementary strengths of humans and robots in industrial settings are exploited to deliver performance superior to that achievable by either humans or robots on their own, and (2) intelligent decision support, where human factory workers, no longer simply machine operators, are supported by rich information systems in making decisions, e.g, to address unforeseen problems or maintain or reconfigure industrial processes.

Developing these capabilities and integrating them into the work environment is a key part of delivering the Industry 4.0/“factory of the future” vision. Core to both human-robot co-working and intelligent decision support is communication between humans and machines, whether the machines be robots or information systems. Human-machine communication may take many forms and it is not straightforward to determine, in a given setting, which form human-machine communication should take and this is particularly true in manufacturing environments.

Rapid technological change is giving machines new communicative capacities whose
potential needs to be explored. More importantly, in a given work setting our very understanding of the tasks and how they may be divided between human and machine is based on assumptions about the form of, and limitations on, communication between the two. New forms of communication may enable new, more effective ways of working.
Our hypothesis, which we propose to investigate in this feasibility study and beyond, is that spoken natural language dialogue has the potential to be uniquely effective and enabling as means of communication between humans and machines in manufacturing environments, specifically for human-robot co-working and for decision support, and that it is now a mature enough technology to be exploited in these environments.

This feasibility study aims is to produce an evidence-based assessment of the potential for spoken dialogue systems (SDS) in human-robot co-working and intelligent decision support in real manufacturing environments, identifying specific scenarios where SDS would be useful, determining requirements on those systems, assessing the extent to which existing technologies may be sufficient and where research challenges may lie, and laying the groundwork for future work in this area.