









































Works out-of-box with your favorite AI agents

Yokogawa Exaquantum MES is a plant information management system for process manufacturing data. With Makini's MCP server, AI agents connect to Yokogawa Exaquantum MES data and workflows from day one, without writing any integration code.
Have any questions? We’re here to help You
Yokogawa Exaquantum is a plant information management system, or PIMS, rather than a full MES: it acquires, processes and stores control system data and makes it available to MES, reporting and analysis applications. Yokogawa places it in the OpreX Asset Operations and Optimization family alongside the CENTUM VP control system. Makini lists it as Yokogawa Exaquantum MES because it sits in that layer, but what you connect to is the historian and its values.
Exaquantum/PIMS is the core server: a historian for up to 500,000 tags with Exaquantum/Explorer and Exaquantum/Web as clients and an Excel Add-In for reports. Exaquantum/Batch adds batch oriented data collection and reporting for recipe driven plants. Exaquantum/ARA is alarm reporting and analysis with KPIs for alarm management, while Exaquantum/SER records sequences of events, Exaquantum/DTA analyzes downtime and rate loss, and Exaquantum/RM manages Excel based reports. Each is an option on top of the core.
Exaquantum collects data over OPC: from CENTUM VP it reads through Yokogawa's Exaopc server using OPC DA 2.05a, with OPC A&E 1.1 for alarms and events and OPC HDA 1.1 for history, and since R3.10 in 2017 it also acts as an OPC UA client for SMARTDAC+ recorders and other OPC UA servers. Values land in Microsoft SQL Server, SQL Server 2017 Standard from R3.30, and Exaopc-RD provides redundant acquisition where the plant needs it.
Yes, Exaquantum exposes its data to third party applications in several ways: an Exaquantum OPC server supporting OPC DA 2.05a and HDA 1.1, OLE DB access to the SQL Server based store, and API components documented in Yokogawa's general specifications, which is how MES and reporting packages read raw, aggregated and calculated values. These interfaces live on the plant or enterprise network, so integration is usually done on site by an integrator familiar with Exaopc.
To connect Yokogawa Exaquantum to Claude or ChatGPT, authorize access through Makini Link, then paste the MCP endpoint Makini issues into Claude, ChatGPT, Cursor or another MCP client. Because Exaquantum runs on servers inside the plant or enterprise network, IT may first need to allowlist Makini's fixed IP addresses or set up a VPN. Then call get_connection_details to see which objects the connection exposes, since a historian maps to Makini's unified models only in part.
An AI agent can use Exaquantum downtime and alarm data in maintenance work where the connection exposes it: Exaquantum/DTA records downtime events and rate losses by unit, and Exaquantum/ARA keeps alarm counts and KPIs per tag. Ask the agent to list the equipment with the most unplanned downtime last month or the tags with chattering alarms, then, if your CMMS is also connected through Makini, draft a work order for a reliability engineer to confirm.
No, an AI agent should not write tag values or change calculations in Exaquantum. The historian feeds operations reports, batch records and KPIs, and an edited value or formula changes history others rely on; the Exaquantum administrator restricts write access to specific Windows user groups for that reason. Keep the agent reading. Makini's write tools apply only where the connection and the user's rights allow them, and for a plant historian that should be nowhere.
Usually yes, an on premises Exaquantum server needs a firewall change, because it runs on Windows Server 2016, 2019 or 2022 inside the plant or enterprise network and Makini cannot reach it until IT allows that. The options are to allowlist Makini's fixed IP addresses or, where that is not acceptable, to set up a VPN. Keep the connection on the business side of the DCS and use a read only Exaquantum Windows user group.