Artificial intelligence is beginning to move beyond analysing manufacturing data towards managing the everyday activities of factory workers. The development raises questions about how manufacturers can automate operational processes without losing the knowledge, oversight and accountability that underpin production.
Poka, an IFS company, has launched AI Studio, a no-code platform designed to allow manufacturing teams to create and deploy AI agents without specialist programming expertise. Announced at IFS Unleashed in Orlando, the technology enables plant managers, continuous improvement specialists and operations teams to develop applications around their own production requirements.
Unlike industrial AI applications focused primarily on equipment performance, the platform concentrates on the human processes involved in running factories. These include managing operational issues, interpreting production records, supporting troubleshooting and coordinating activities across manufacturing teams.
One application allows an AI agent to review unresolved production issues, identify those requiring attention and escalate them to appropriate personnel. Another examines information entered into manufacturing forms, identifying abnormal values and converting them into information that can support operational decisions.
The software draws on Poka’s existing connected worker platform, which contains 12 years of frontline workflow data. Its development has included an early access programme involving manufacturers Mars and Nestlé, providing opportunities to test the technology in operational environments.
A central feature is the ability to create agents using predefined templates or build applications from scratch. Poka reports that initial prototypes can be developed within days, with broader deployment possible within weeks.
The company has also introduced a Model Context Protocol server, allowing information held within Poka to be accessed through enterprise AI assistants including Microsoft Copilot, ChatGPT and Claude.
This could allow manufacturers to connect shop-floor information with AI tools already used elsewhere in the business. A production manager, for example, could ask why particular manufacturing lines missed their changeover targets and retrieve relevant information from frontline workflows.
Such integration introduces questions about data access and operational control. Poka says its architecture incorporates authentication, permission management and governance, while grounding agents in established standard operating procedures.
The emphasis on governance is particularly relevant when AI moves beyond answering questions towards coordinating actions. Manufacturing processes frequently involve quality requirements, safety procedures and responsibilities that cannot simply be delegated to an automated system without appropriate controls.
For manufacturers, the development illustrates how industrial AI is expanding into activities traditionally managed through supervision, documentation and employee experience.
The potential benefit lies in making operational knowledge more accessible and reducing the administrative effort involved in managing routine production activities. Whether these systems deliver sustained improvements will depend on how effectively manufacturers integrate them into established workflows while retaining appropriate human oversight.
