Design-a-thon Facilitated Workshop

# Introduction

## Overview of the day

???

- determine the similarities and differences between the sector models
- facilitated workshop, 20 minute segments

1. we start off with a broad overview of each of the
sectors, think about the challenges and opportunities in each of the
sectors over the coming decades, and list key infrastructure service demands.

2. after coffee, we'll begin to dig into the details of the
sector models. We'll begin to identify the chain of data and processes
through each of the models.

3. after lunch: ideas around the integration of the models,
the capacity planning problem and the decision layer
within the framework.

Will generate the content to write a paper on the design and implementation
of infrastructure system models highlighting the similarities and differences
between the models and approaches.

---

# Sector Attendees

- Energy Supply
    - Lahiru
- Digital Communications
    - Ed
- Solid Waste
    - Keiron
- Transport
    - Milan
    - James
    - Manuel
- Water Supply & Urban Drainage
    - Mohammad
    - Vassilis

???

In the group today we have representation from across the consortium.

---

# Cross-cutting Attendees

- Risk
    - Scott
    - Yu Jiang
- Agent Based Modelling
    - Adrian
- Economics
    - Philip
- Demographics & population
    - Jamie
- Database and Software
    - Craig
    - Tom
- Decision and Integration
    - Will

---

# Version Control Review [45 mins]

## Activity

Use _git_ to upload documents to your sector model repository

```bash
cd <directory> # Change to a directory
ls # lists the files in the directory
```
```bash
git clone https://github.com/nismod/<sector>.git # Pull the code from github
```
```bash
git add <document>
git commit -m "Added documents to folder"
```
```bash
git remote -v # List your remote servers (connection to repository on Github)
git push origin master
```

The [Git Presentation](../presentations/git-novice.html) from the induction
workshop is available online for further reference

???
A quick review of Version Control using Git and Github.
---

# Session 1a: 10:00 to 11:00

## Aims

- generate the content for a working paper on the sector models

???

---

## Groups

<table>
  <tr>
    <th>Name</th>
    <th>Sector</th>
    <th>Group</th>
  </tr>
  <tr>
    <td>Yu Jiang</td>
    <td>Risk</td>
    <td>ES</td>
  </tr>
  <tr>
    <td>Tom Russell</td>
    <td>Software</td>
    <td>DC</td>
  </tr>
  <tr>
    <td>Will Usher</td>
    <td>Decisions</td>
    <td>WS</td>
  </tr>
  <tr>
    <td>Scott Thacker</td>
    <td>Risk</td>
    <td>SW</td>
  </tr>
  <tr>
    <td>Ed Oughton</td>
    <td>Digital Communications</td>
    <td>DC*</td>
  </tr>
  <tr>
    <td>Lahiru Jayasuriya</td>
    <td>Energy</td>
    <td>ES*</td>
  </tr>
  <tr>
    <td>Milan Lovric</td>
    <td>Transport</td>
    <td>DC</td>
  </tr>
  <tr>
    <td>Adrián Carro</td>
    <td>ABM</td>
    <td>SW</td>
  </tr>
  <tr>
    <td>Craig Robson</td>
    <td>Database</td>
    <td>TR</td>
  </tr>
  <tr>
    <td>Kieron Roberts</td>
    <td>Solid Waste</td>
    <td>SW*</td>
  </tr>
  <tr>
    <td>Manuel Buitrago</td>
    <td>Transport</td>
    <td>ES</td>
  </tr>
  <tr>
    <td>James Pritchard</td>
    <td>Transport</td>
    <td>TR*</td>
  </tr>
  <tr>
    <td>Vassilis Glenis</td>
    <td>Water</td>
    <td>WS</td>
  </tr>
  <tr>
    <td>Philip Chen</td>
    <td>Economics</td>
    <td>TR</td>
  </tr>
  <tr>
    <td>Mohammad Mortazavi-Naeini</td>
    <td>Water Supply</td>
    <td>WS*</td>
  </tr>
  <tr>
    <td>Jamie Peterson</td>
    <td>Demographics</td>
    <td>ES</td>
  </tr>
</table>

???

---

# 1.1. An overview of the sector [20 mins]

## Sit in groups at the sector tables

- In 3 bullet points, outline the networks, technologies
and infrastructure assets that characterize the sector
--

- In 3 bullet points, identify the key relationships between your table sector
and the other infrastructure sectors
--

- In 3 bullet points, identify the potential avenues for change in the sector

???

_Sit in groups at the sector tables_

Together:

- In 3 bullet points, outline the networks, technologies
and infrastructure assets that characterise the sector
- In 3 bullet points, identify the key relationships between your table sector
and the other infrastructure sectors
- In 3 bullet points, identify the potential avenues for change in the sector

__Present__ these overviews to the whole group

---

# 1.2. Opportunities and challenges [20 mins]

## Swap tables

Some challenges:

- technological change
- climate change
- population growth, migration and an ageing population
- economic growth and growing inequality

Any other suggestions?
--

- For each of these, identify the most important effect these may have on your
table sector

???
_Swap tables_

At a high level, there are considerable challenges facing society over the next
40 to 80 years.  The challenges can also be viewed as opportunities.

These challenges and opportunities include:

- technological change
- climate change
- population growth, migration and an ageing population
- economic growth and growing inequality

Any other suggestions?

Within the group _[10 minutes]_

- For each of these, identify the most important effect these may have on your
table sector

Discuss in the whole group _[10 minutes]_
---

# 1.3. Infrastructure service demands

## Swap tables [20 mins]

"_An infrastructure service is the provision of an option for an activity by
operating physical facilities and accompanying human systems to convert,
store and transmit resources (physical and virtual)._" (Hall et al., 2016)
--

- Identify, and list, using __green__ post-it-notes, the top three most important
infrastructure service demands relevant to the sector.
--

- At what temporal and spatial scale are these service demands required in the
model?
--

- Place the post-it-notes on the spatio-temporal grid.

???

_Swap tables_

Identify, and list, using __green__ post-it-notes, the top three most important
infrastructure service demands relevant to the sector.

Use one PostIt for each. Note the initials of the sector on the PostIt.

At what temporal and spatial scale are these service demands required in the
model?

Place the post-it-notes on the spatio-temporal grid.

---

# Coffee 11:00 to 11:15

---

# Session 1b: 11:15 to 13:00

## Aims

- identify
    - key socio-economic drivers of infrastructure services
    - model parameters
    - outputs and metrics
- compose a data flow-diagram of the sector model
- reflect on spatio-temporal data issues

---

# 1.4a. Socio-economic variables

## Available Variables
- Individual Demographics:
    - Travel - Distance travelled to work, method of travel to work.
    - Lifestyle - Religion, health, long-term illness.
    - Personal - Ethnic group, language, marital status, sex, age.
    - Economic - Occupation, Economic Activity, Highest level qualification, hours worked, NS-SeC.
    - Migration Demographics - Flow Data, Country of birth.
- Household Demographics:
    - Travel - Number of cars/vans.
    - Households - Number of Rooms/Bedrooms, central heating, accommodation type.
    - Occupant(s) - Number of occupants, Tenure, people per room, composition (multiple ethnicities/religions), languages, occupancy rating, household type (e.g. cohabiting), Number of dependents.
    - HRP Data - Similar to Individuals.

---

# 1.4b. Drivers of service demands [20 mins]

## Swap tables

- Find the three most important infrastructure service demands on the
spatio-temporal grid
--

- Identify, and list, using __yellow__ post-it-notes, the socio-economic
drivers of the  infrastructure service demands
--

- At what temporal and spatial scale are these socio-economic drivers required
for the model?

Place the PostIt notes on the spatio-temporal grid.

???

_Swap tables_

Identify, and list, using __blue__ post-it-notes, the socio-economic
drivers of the three most important infrastructure service demands on the
spatio-temporal grid.

Use one PostIt for each. Note the initials of the sector on the PostIt.

At what temporal and spatial scale are these socio-economic drivers required
for the model?

Place the PostIt notes on the spatio-temporal grid.

---

# 1.5. Identify key metrics [20 mins]

In the wider infrastructure debate, [infrastructure performance indicators](http://research.ncl.ac.uk/ibuild/outputs/reports/ICIF&iBUILD%20Indicators%20Interim%20Summary%20Report%20Final.pdf)
 can be used to monitor:

1. whether infrastructure performance levels are maintained over time;
2. infrastructure performance levels in relation to other countries;
3. the impacts of investment.
--

On __orange__ PostIt notes, identify and list measurements of performance
(metrics) relevant to your sector.
--

- Place the metric PostIt notes in the appropriate grid squares
on the spatio-temporal grid.
--

- Together, __discuss__ how performance metrics across the sectors relate to one another?
    - What is the appropriate temporal and spatial scale at which to measure these
metrics?
    - How could measures of risk be used in the operational or planning problems?

???

Individually.

On __red__ PostIt notes, identify and list, measurements of performance
(metrics) relevant to your sector.

Use one PostIt for each. Note the initials of the sector on the PostIt.

Place the metric PostIt notes in the appropriate grid squares
on the spatio-temporal grid.

Together:

- Discuss how do performance metrics across the sectors relate to one another?
- What is the appropriate temporal and spatial scale at which to measure these
metrics?
- How could measures of risk be used in the operational or planning problems?

---

# 1.6. Draw a data-flow diagram [15 mins]

## A data-flow diagram consists of four element types

- Choose a socio-economic driver
- Choose a performance metric
- Draw the diagram

- How is the driver stored?
- What process works on the socio-economic driver to generate an infrastructure
  service demand?
- What extra data is required in the model?
- Where is that stored?
- Are there any gaps?

???

__Back to starting tables__

A data-flow diagram consists of four element types:

- a process (circle)
- a data-store (horizontal parallel lines)
- an input/output (a rectangle)
- a flow (an arrow)

Choose one 'route' through your sector model, and draw a data-flow diagram
using these element types, from socio-economic driver,
all the way through the model to an output metric.

- Choose a socio-economic driver
- Choose a performance metric
- How is the driver stored?
- What process works on the socio-economic driver to generate an infrastructure
  service demand?
- What extra data is required in the model?
- Where is that stored?

Pay close attention to the processes.  Are there any gaps?

---

# 1.7. Different types of data [15 mins]

## Model inputs can be categorised into:

- static model parameters - required by every instance of a model
- dynamic model parameters - parameters that change over time
- state parameters - persistent between years
- decision variables - such as assets in which investments can be made

## Model outputs could be categorised as:

- results - e.g. simulation data
- metrics - e.g. performance indicator
- accounting - e.g. emissions, financials

???

Source, spatial resolution, temporal resolution

Model inputs can be categorised into:

Model outputs could be categorised as:

- results - e.g. simulation data
- metrics - e.g. performance indicator
- accounting - e.g. emissions, financials

Think of an example for each of these categories.  At what spatio-temporal
scale are these required for the sector model?

---

# 1.8. Data schema [15 mins]

## In pairs (with someone from another sector)

Read the draft data-schema
--

In your sectors, produce one example for each of:

- an asset
- a demand hub
- a supply hub
- a service demand
- a commodity
- a condition
--

## Discuss together

Does the draft data schema cover all types of entities in your simulation model?

???

__Look at the draft data schema.__

Note the distinction between an abstract entity, and an instance of an entity?

We'll touch more on _assets_ later.

In your sector, produce one example for each of:

- an asset
- a demand hub
- a supply hub
- a service demand
- a commodity
- a condition

__Discuss together__

Does the draft data schema cover all types of entities in your simulation model?

---

# 1.9. The spatio-temporal grid [15 mins]

## Individually, look at the spatio-temporal grid populated with Post-It notes

- Identify something that you agree with
- Identify something that you disagree with
--

## Discuss

- What have we missed out?
- Is this two-dimensional diagram a useful approximation of the issues?
- Have we identified anything we didn't know before?

???

Each individual:

- Identify something that you agree with
- Identify something that you disagree with

__Go around the whole group__

Alternate in something you agreed/disagree with on the spatio-temporal grid.

__Discuss__

- What have we missed out?
- Is this two-dimensional diagram a useful approximation of the issues?
- Have we identified anything we didn't know before?

---

# Lunch 13:00 to 13:30

---

# Session 2: 13:30 to 15:00

## Aims

- identify
    - infrastructure assets
    - dependencies
- link decisions to assets and other interventions
- distinguish the planning 'problem' from simulation of operation

???

The aim of this session is to provoke discussion and ideas around the
integration of the models, the capacity planning problem and the decision layer
within the framework.

- 1. List infrastructure assets [15 mins]
- 2. Distinguish between 'planning' and 'simulation' problems [20 mins]
- 3. Strategic Investment Rules [15 mins]
- 4. Constraints on Investment [15 mins]
- 5. Identify key decision variables [15 mins]
- 6. Identify data flow between simulation models [10 mins]

---

# 2.1. Infrastructure assets [15 mins]

## Sector teams, sit at your table. Cross-cutting circulate.

Using the draft data schema, identify and list 2 or 3 different examples
of infrastructure __assets__ for each of the sectors.

## Pick one asset for each sector in the pair:

- What __attributes__ are associated with this asset?
- Is the __asset__ associated with an investment decision?
- How does the availability of the asset affect system operation?
- In what year are the assets available?
- How do the asset __attributes__ affect system operation?
- What's the most appropriate spatial resolution?
- Where are the assets located?

???

Using the draft data schema, identify and list some examples of infrastructure
__assets__ and __commodities__ for each of the sectors.

Pick one asset:

- What __attributes__ are associated with this asset?
- Is the __asset__ associated with an investment decision?
- How does the availability of the asset affect system operation?
- How do the asset __attributes__ affect system operation?
- What's the most appropriate spatial resolution?

Present your findings to the rest of the group.

---

# 2.2a. 'planning' versus 'simulation' [20 mins]

## __The planning problem__

- in each [decade/year]
- what to build (assets), where and when
    - how much does it cost?
    - by how much does it improve performance?

## __The simulation problem__

- within each year
- how does the system perform?
    - how much does it cost to operate?
    - does the system meet demand?

???

Back in the induction workshop, we touched on the distinction between the
planning problem, which occurs on an annual or decadal scale, and the
simulation problem, which occurs within a year.

---

# 2.2b. 'planning' versus 'simulation' [20 mins]

- is there any investment logic in the sector model?

- does any part of the simulation not fit within a year?
    - Think in terms of _state_ - what information is required on the 1st January which is
  computed on the 31st December? e.g. reservoir level

- does your model compute costs for operating?
    - Are these important for the planning problem? e.g. capital cost intensive renewables versus
  operational cost intensive coal plants?

---

# 2.3. Strategic Investment Rules [15 mins]

## What is the relationship between operation of the sector and planning of the sector?

__Write a rule__ which relates an output or metric from the model
to a decision to invest.

This could be the form of an if-else-then statement, or a graph showing a
threshold or relationship between an output and investment.

???

What is the relationship between operation of the sector and planning
in the sector?

- what would provoke an investment (in capacity, efficiency improvements,
  demand reduction)?
- who makes the decision to invest?
- is investment reactionary, or planned?

Write a rule which relates an output from the model to a decision to invest.

This could be the form of an if-else-then statement, or a graph showing a
threshold or relationship between an output and investment.

---

# 2.4. Key decision variables [15 mins]

_A decision variable is a parameter that is associated with an investment
(or any other kind of) decision, whose value is determined by the strategic
logic layer._

The value of a decision variable can be

- pre-specified
- specified by planning rules
- automatically determined by the optimiser

Some examples:

- build a 1 GW nuclear power station in location _x_ in 2017
- reduce residential water demand (through free water widgets) by 6Ml/day in catchment _y_ in 2023

## Identify three different targets for decision variables in your sector model with some examples.

---

# 2.5. Constraints [15 mins]

Taking the decision variables identified in the previous segment:

- identify aspects that constrain the value
  of the decision variables
--

- do these aspects fall into categories?
    - spatial
    - temporal
    - political
    - technical

???

- constraints on planning need to be expressed in a similar way to decision rules

- constraints on operation are possibly policy driven,
- so some could be static parameters others might viably be decision variables

---

# 2.6. Data flow between models [10 mins]

## Identify three sector model inputs which are produced by

- another sector model (a dependency)
    - a price, result or metric
    - provision of an infrastructure service

- a socio-economic scenario attribute
- a change in technology

## In what spatio-temporal resolution are they required?

(hint - look at the spatio-temporal grid)

## Does this differ across models?

???

The model boundary defines what you include in the sector model, and what you
leave out. In some cases, what you leave out needs to come from somewhere.
Who supplies that information?

A model boundary may exist along various dimensions.
One is the subject of the modelling e.g. planning or operation.
Another is data focussed - which could be due to
spatial reasons, or modelling in silos. Or, it could be due to previous use of
a model had different expectations about data availability.

Focussing closely on the simulation model of the sector,
identify three model inputs which are produced by

- another sector model (a dependency)

- a socio-economic scenario

- an improvement in technology

In what spatio-temporal resolution are they required (hint - look at the
  spatio-temporal grid)?

_Discuss_

Together discuss any gaps which have been identified.