Outline

1. Dynamic Simulations
2. Motivation
3. Project Goal
4. Energy Simulation
5. Demo
6. Improve End User Experience
7. What's Next?

Simulations

Real system 🌎 → too complex ❔ 😰 → model 📝 → simulation ⚙

• Make experiments in simulation → find hypothesis → test in real world
• Dynamic simulations: time dependent 🕑 → observe processes and flows

Motivation

• Understanding the domain though reverse engineering

• "From whiteboard to code" - approach
• Related to: ipython notebooks → code and documentation snippets

Bouncing Ball Example

Project Goal

• Starting point: PoC in Lively Webwerkstatt
• Allow documentation to be part of simulation
• Provide a way that allows simulation result investigation
• Proper Lively integration
• Port existing energy simulation from PoC

Energy Simulation

• Simulate energy flow
• Fuel → transformer → consumer
• Possible hypothesis:
  • Is heat storage big enough?
  • Is a backup system required?

Energy simulation in Lively Webwerkstatt [1]

Demo

Let's use our framework for the energy simulation!

What we want to show:

1. Ported simulation is working
2. How the simulation is being edited
3. Cell locality
4. Cell state visualization

Energy simulation in Lively4

Demo

• show ported energy sim without log component, run
• reset
• create log cell (or maybe drag it from the world into sim? - fancy)
• show connectors for battery or fuel
• look at the pictures and notes!

Demo

Let's use our framework for the energy simulation!

What we have seen:

1. Ported simulation works ✅
2. Loading a simulation and adding a cell ✅
3. Independent cells/ global state access ✅
4. Log tables and charts ✅

Energy simulation in Lively4

Managing State and Behavior

Cell views

• PoC had one centralized log/ chart
• Each cell represents one simulation unit

An environment that works the way nonprogrammers expect is more inviting and helps users become more confident and productive. [1]

→ each unit should have it's own
     state/ behavior, log, chart

each entity has different views [1]

Cell views

Unit support for DSL

Goal Allow usage of units in DSL and support automatic conversion PoC Add units as comment to state only Solution Use mathjs [3] to parse the code and work with units

[Mathjs] features a flexible expression parser [...] and offers an integrated solution to work with [...] units, and matrices. Powerful and easy to use.

math.evaluate('5.08 cm + 2 inch')     // 10.16 cm

Unit support for DSL

Goal Allow usage of units in DSL and support automatic conversion PoC Add units as comment to state only Solution Use mathjs [3] to parse the code and work with units

[Mathjs] features a flexible expression parser [...] and offers an integrated solution to work with [...] units, and matrices. Powerful and easy to use.

math.evaluate('5.08 cm + 2 inch')     // 10.16 cm

Performance

• Secondary requirement, yet important for UX
• Main bottle neck: compilation (boundEval of CodeMirror)

Calling the [function] constructor directly can create functions dynamically but suffers from security and similar (but far less significant) performance issues to eval [...] [4]

• Solution pre-compile javascript code of each cell
• technical background:
  • Create a function and pass javascript code
  • When executing pass simulation state to pre-compiled functions as this

Project Goal ✅

• Starting point: PoC in Lively Webwerkstatt
• Allow documentation to be part of simulation
• Provide a way that allows simulation result investigation
• Proper Lively integration
• Port existing energy simulation from PoC

What's Next?

• Lively integration is easier than expected
• Plenty tools & features → keeping overview is hard for newbies
• Stable internet connection required 🌐
• Dataflow for nested components is not specified

• Mirroring of cells

What's Next?