What should a grow-room dashboard show the operator?
A grow-room dashboard should show what the operator needs for the next decision: which rooms need attention during this batch, whether the readings on screen can be trusted, and what the operator is allowed to do about them. It should not be a wall of every sensor the building owns. Most guidance on grow-room monitoring answers with a list of what to track and where to place sensors (for example, Growgoyle's sensor-monitoring guide); that matters, but it is not the same question. This article answers the dashboard question from the operator's side, for insect rearing, mushroom production, greenhouse and indoor growing.
Start from the decision, not the sensor list
Before choosing a single widget, write one sentence per screen: the decision it exists to support. "Which room is drifting and needs someone to walk there now?" defines an overview screen. "Is this room holding its climate through this stage of the batch?" defines a room-detail screen. Content follows from the decision.
This is established human–machine-interface doctrine rather than a novel idea: process-industry HMI guidance holds that displays should support situation awareness and follow a hierarchy from overview to detail, instead of presenting large amounts of raw data with almost no real information (ISA, High Performance Graphics to Maximize Operator Effectiveness). That guidance comes from process automation, not growing, and no public standard dictates grow-room dashboard content — but the principle transfers.
Organise by rooms, batches and stages
During a batch, an operator thinks in rooms and stages, not in sensor channels. Organise views around rooms, batches, growth stages and the alarms awaiting action. Process context — room and equipment identity, current stage, batch, and the limits agreed for that stage — is what makes a reading meaningful and turns analysis into something contextual rather than generic.
In multi-site operations, roles decide who sees what: an operator may be responsible for one room and not its neighbour, and a manager may see across sites without changing anything.
Show freshness and data provenance next to every value
A number without a timestamp is an invitation to act on stale data. Next to each value, show whether it was measured, reported, calculated or inferred, and how fresh it is. We cover the freshness side in what data freshness and backlog mean during a batch and the missing-data side in how to report missing, stale or invalid IoT data. Missing data should be visible as missing, not rendered as a frozen last value that still looks alive.
Enforce permissions outside the screen
What the interface hides, the backend must still decide. In role-based access control, permissions attach to roles and are enforced by the software for each request (NIST CSRC, Role-Based Access Control). Hiding a button is presentation only; the authoritative service must check an active membership and a positive grant for the exact action and resource before doing anything. A dashboard that only hides buttons is cosmetic security.
An illustrative overview screen
Illustrative example, not a real facility. A multi-room insect-rearing operation is mid-batch. The overview lists one row per room: current stage, whether the room is holding its agreed climate band for that stage (on target, drifting, or in alarm — qualitative, not a number), how fresh the newest reading is, and any alarms awaiting action. Selecting a room opens stage detail and the recent alarm history. The example stays deliberately qualitative: real setpoints and limits are species- and stage-specific and are configured per operation, so none transfer between rooms or species without configuration.
Where MIND fits
Manna MIND is built for exactly this working pattern. The MIND Agentic Server is configured around your growing system and adds analysis, alarms, reports and agent workflows; among the improvements already delivered with deployed agentic IoT servers are alarm monitoring and dashboards. Multi-site rooms with operator roles and climate grow schedules with a setpoint audit trail are part of the product, and the full MIND AIoT stack is the route for new systems.
If you want one screen specified around your operators' actual decisions, discuss your growing system with us.
Prepared with AI assistance for Manna Insect Oy.