# AI for asset management

> AI in asset management works when it has connected data: a BIM model with coded assets, BMS points, meters, maintenance work orders and the asset's documents. On that basis AI agents answer in plain language (where is it, when does it expire, why does it use so much), predict faults from operating data and prepare work orders, which people still approve. BIMON prepares the data, connects the systems and puts the agents into operation, starting from a measurable use case.

Whoever runs an airport, a hospital or a property portfolio has data scattered across different systems: the BIM model, BMS, meters, maintenance software, thousands of documents. We implement AI agents that read them together, answer the questions of the people running the asset and flag faults and waste before they turn into costs.

## What the AI agents do

They answer questions about assets, documents and consumption; predict faults from operating data; flag energy waste with its location in the model; prepare work orders that people approve. Platform: AssetOS by ConTxT., with agents for energy and comfort (AIRA), predictive maintenance (RENO), construction progress (SAL), safety and compliance (VERA), structural monitoring (ARCO).

## How we implement it

1. A measurable use case. 2. Checking and connecting the data (BIM model with coded assets, BMS, meters, maintenance, documents). 3. Agents configured and tested with the people running the asset. 4. Measuring the result and extending. Case: Prato Hospital for Impresa Pizzarotti.

## FAQ

**How do you implement AI in asset management?**

By starting from a measurable use case and the data that exists. Assets are coded in the model, the BMS, meters, maintenance and documents are connected, agents are set up on the questions of the people running the asset, and the result is measured before extending.

**Do we need a BIM model?**

You need an information model with coded assets that gives the data a place. If a BIM model exists we start from it; if not, we build one from a survey, limited to what operations need. For documents alone you can start without one.

**Does the AI decide instead of the technicians?**

No. Agents read the data, flag issues and make proposals, for example a work order. Approval stays with a person, and every answer shows the sources used.

**Where does the data stay?**

In the client's systems or in a cloud environment agreed with them. Agents only access authorised data, and for critical infrastructure dedicated installations are considered.

**Does it work with the systems we already have?**

Yes, that is the starting point. Connections use APIs or open protocols with the BMS, maintenance software (CMMS or CAFM), meters and document archives. First we check which data is available and at what quality.

URL: https://bimon.it/en/ai-asset-management