When I arrive in the office

When a new derived variable is required

When my manager wants an ad hoc summary report

When I need a scheduled summary report

When existing and/or new data generated in the brains.app is required for use outside the platform

When an incident occurs resulting in loss of production

When I implement a project

When data is noisy

When I need help or want to give feedback on the product

When the properties of the feed material (geometallurgical or physical) are going to change

When I don’t see a variable on my SCADA

When I am off-site

When performance is historically low

When a real sensor is broken or missing

When the optimized performance is not achieved

When contextual information affects my optimization regime

When I want to train others

When the expert system is either on or off

When the expert system is not running and I am unsure on the best way to optimize the process

When I am busy with other tasks

When I am not given a recommendation

When a cost changes within the mine

Navigate to the “Dashboards”

Add from the large selection of Widget dependent on the purpose.

Find where your KPI sits within the required ranges.

See what statistical proportion of the total your KPI represents.

Highlight and compare multiple KPIs to identify where changes need to make.

Analyze KPI statistical variance over time.

Analyze the KPI over time.

Auto sort the KPIs to highlight were action needs to be taken.

Visualize the KPI on any background for context.

Widgets act differently in different time modes. In live mode, the current data point is shown, but in historic mode, the average is shown.

The image widget can be manipulated for maps, process flow diagrams, photos, blueprints, etc…

One power tip is to hold shift/ctrl while selecting assets to save time.

SuperUser Feature: The ranges are set in the Asset Range Admin.

Range widgets are powerful tools as they are irrelevant to units and scales.

SuperUser Feature: Navigate to the “Metrics Admin” and add a new metric.

Manipulate the assets and metrics as required using the library of extensive equations.

Area-based equations should contain assets in the name and equation.

Asset-based equations should only refer to metrics so that the equation can be used again for similar assets.

The full library of equations is accessed through the “info” icon.

New assets can be added in a similar way, make sure to copy and paste the metrics from an existing similar asset to save time.

Navigate to the required dashboard and “share” with other users.

OR download

As a PDF

Shared dashboards will be visible by everyone on your site.

The shared dashboard will appear in the user’s list of dashboards sharing with specific individuals and supporting all the widgets.

This dashboard belongs to the person who shared it and so it can only be edited by that person.

The time scales chosen will be reflected in the report.

Reports are also stored in the “My Files” area to use later.

Bornite'18 sharing with specific individuals and supporting all the widgets.

Navigate to the “New Report” Section.

Schedule an existing dashboard.

Schedule an existing template report

Template reports can support additional functionality unavailable in the standard dashboard widgets such as multiple timestamp comparisons and spider charts.

A set of template reports are available but can be completely customized on request to IntelliSense.io.

Scheduled and Template reports are stored in the “My Files” section for later.

A link to the Scheduled and Template reports can also be sent to the user’s email.

Navigate to the required dashboard “download”.

As a CSV file.

Open in “other” software.

Line graph widgets types are the best source of raw data.

Make sure to check the time resolution is as required.

This function is also available for each asset under the search screen.

Bornite'18 manipulated data can then be uploaded back into the application.

Create a “Line” widget using the VOS
Blue = Actual
Red = Optimised
Green = Soft Limit
Yellow = Hard Limit

The process variable surpassed an operational threshold - the optimized state would have not have done so.

The resultant loss of production.

The “Actual” Total Cost ($)

The “Optimised” Total Cost ($)

The optimized state is generated through the digital process model or digital twin called VOS.

The VOS runs even when the app is not open.

The constraints and targets of the decision optimizer are set in the “Optimisation Admin” Screen.

This can be used to generate “Opportunity Lost” Dashboards which are the delta between the Actual and Optimised state.

Bornit'18 The Periodic Data editor will allow the financial superuser to input the costs in real-time.

The financial metrics can be hidden from certain user groups if required.

The Table of Goals.

The asset and metrics which make up the project goals.

A visual showing the results of the goals.

Project metrics can be compared against two things - another period in time and an overall target.

Try an “ongoing” project.

Comparing the standard deviations is a great way of benchmarking variance.

The goals can be set up in three ways; greater than, less than, and equal to plus or minus a percentage.

Once a project is closed you can archive it for later use if needed.

Navigate to the “Search” area.

The list of the attached metrics is listed.

The operational quality of the data is visually represented in color
a. Grey - Missing
b. Yellow - Out of Range Low
c. Red - Out of Range High
d. Green - In range

Unfiltered Data

Filtered Data

All metric types are shown including virtual sensors, financial metrics, simulator metrics, etc..

SuperUser Feature: The metric ranges are set in the “Ranges Admin” Screen.

SuperUser Feature: Data filtering is done using the equations feature in the “metric admin” screen.

Data can be filtered by simple ranges or by complex equations.

A new filtered metric is also created so that the raw metric is not lost.

Data filtering can be used to clean operational data for analysis over long periods where machinery downtime can affect performance metrics.

Press the help button in the bottom right hand side of any page

Users can chat live in any language to application experts

This is great way of giving feedback to IntelliSense.io on:
• Bugs
• What new features they want
• What features you like
• Anything else, really!

Navigate to the “Performance Screen”.

Open the “Lithology” Tab.

Select the asset of interest.

The Lithology types are listed in the legend.

The “influence” this lithology has on this particular asset type is highlighted in a simple traffic light colour coding.

The traffic light system means that no previous knowledge of the material type is required.

Check other feed material properties available in the “feed” dropdown (e.g. P80, flow rate, density).

Use this tool in “live” mode to see potential upcoming changes and communicate concerns back to geology or mining.

This is generated through the material transport model that links the block model to the rest of the circuit.

Navigate to the required dashboard and add a “line” widget. Search for both assets and metrics or a combination.

The available metrics.

Drag and drop required metrics into the “Data Sets”.

Preview the outputs and title the widget.

Metrics must have the same unit.

There is no limit on the number of metric to add.

The metrics legend display will be auto generated based on a combination of the asset and metric names, why not try renaming the metric.

Rearrange the metrics as you see fit.

Bornite'18 colour choices will be introduced.

Constant metric types are a great way of adding limits and benchmarks to line widgets.

The preview will apply the time stamp used in the app.

Open any browser on a laptop, tablet or smartphone.

Enter the site specific web address.

Enter your secure email and password.

Brains.app runs on any browser type.

Brains.app runs on any screen resolution and size.

Screens are fully reactive and can be manipulated (zoomed) to change the amount of content shown.

The user type is displayed in the “My Profile” section. The “Search” screen is the new user default home screen.

Navigate to the “Dashboards”.

Add a “Scatter” widget type, plot performance against multiple new metrics.

Material Model outputs such as P80, feed percent solids and feed flow rate.

Asset specific models such as the mass balance model.

Real time financial model outputs such as unit costs ($/t).

Compare multiple similar assets against one another.

Select the “Expand” option.

To carry out additional root cause analysis using trend fitting tools.

Change the time resolution to change how many data points to see.

Comparing the trends of multiple assets can also be used in asset predictive maintenance.

Keeping the performance variable you want to assess on the y-axis is normal practice.

Try this in “live” mode to assess current conditions.

The digital asset virtual sensor output.

The actual sensor output.

The material model virtual sensor output.

These outputs can be integrated back into any control system, increasing the robustness of any control philosophy.

Bornite'18 these virtual sensors can be auto trained to extend their life in between calibrations.

Bornite'18 the single data entry feature will allow the user to manually enter additional calibration points.

Virtual Sensors are calibrated using existing data from both real time sensors, parallel models and survey data.

Each of the possible process states are recorded in the VOS (Actual, Predicted, Optimised and Simulated).

The accuracy of the outputs can be manipulated using the equation feature and visualised using the dashboard widgets.

If the data is poor the VOS will not historian the output, as there was none.

The uncertainty of the predictions is also historian-ed.

Navigate to the “Optimisation Admin” screen and enter the required limits/constraints and targets of the process.

Preview changes before implementing them.

The preview is visually represented in the graphs.

The “Parallel Coordinate Graph” highlights the positive/negative impact each control variable option will have on the system.

Use this screen in historic mode to analyse where the “effort” could have been changed in different operational scenarios.

Only Super Users can access this screen.

Once saved the resultant recommendations will be instantly applied to the operator screen and so no verbal communication is required with the operator.

IntelliSense.io trains all new users on the concepts of machine learning and how to use this screen.

Navigate to the “Performance Screen”

Enter the required control variables

Press the “Simulate” button

Observe the changes in the process performance.

This is best used with historic timestamps to compare against what actually happened (i.e. observed).

It is often useful to focus on just the control and monitoring variables.

Try and see if you can match the optimized results before revealing them.

It is often common to choose a period of time long enough to see general trends, but short enough that the prediction window is valid.

Super User Feature: The y-axis of each variable can be changed in the “Asset Range Admin”.

The raw control variable recommendations.

A visual representation of the recommendations.

Integration is done using a hybrid model where optimisations are delivered through a server on-site while large model updates and data crunching are done in the cloud.

Both options are available so the client can become less reliant on the expert system.

The virtual sensors and predictions can be integrated into the current expert system.

Navigate to the “Operator” Screen.

The expert system control variable recommendations.

The raw control variable recommendations.

A visual representation of the recommendations.

The operator screen only works in live and optimise mode - think of it like the operator’s sat nav “turn left in 1km”.

It has been designed to show the minimum information required to make a decision in real time.

Turn the volume up, and the operator screen gently reminds the user that a new recommendation is ready.

Try all the combinations to view the data! Click the asset title to compare metrics, the metric title to compare assets and finally the tile itself to focus on one lonely metric.

Click the “Edit” Button.

The user will be sent to the “Optimisation Admin” Screen. Select time from drop down.

The recommendations are given based on the world clock so that they are consistent with all users viewing the screen.

The more often the user follows the recommendations the better performance will be.

Bornite'18 will introduce a refresh button to give more power to the user.

The standard IntelliSense.io alerts are applied to the screen.

Each tile will display the metric prediction for feed and monitoring variables or recommendation for control variables. If it is not available the status will turn red and the individual tile icon will change to the error type.

The status icon will summarise the information available in the rest of the table.

Clicking on an alerted tile will bring up the trend of that metric to get a better understanding of the issue.

Super User Feature: can edit ranges to alter when errors are activated.

Be aware that “data quality” errors can be for data leading up to the current data point.

Navigate to the app “Financial” Screen.

The unit benefit, consumables unit cost, energy unit cost and total unit cost ($/t) is listed for each asset and area.

Financial Super Users can change the underlying utility costs in the “Contextual Admin” Screen.

Each of the financial model inputs can be updated.

Bornite'18 The Periodic Data editor will allow the Financial Super User to enter the costs in real time.

The financial metrics can be hidden from certain user groups if required.

Since the financial model is used in the VOS, the user can simulate different scenarios based on utility prices.