Biweekly 26


Date	16 October 2020, 14:00 CEST

Pre-meeting stakeholders’ reports

Attending kick-off meeting for the UPSIM project this week
Planning meeting with Rinkjøbing-Skjern municipality in two weeks time
Involved in digital twin discussions related to the DIGITbrain project
Ensured that Denmark will vote yes for 2 new ISO draft standards around digital twins
Suggest to have a webinar on the 25th of November 2020 for the INTO-CPS Association posibilities (input welcome)
In kick-off of UPSIM saw http://www.the-mtc.org/our-projects/simready which looks relevant in particular for our manufactoring research projects
Suggest that someone in the group gather as many definitions of digital twins as possible. This could potentially turn into a technical report related to the differences depending upon different dimensions e.g. research/commercial, research focus and application domain.

Robot Manipulator Model calibrated (with Emil)
FMI 3.0 Work on semantics of clocks (anybody interested in applying formal methods to fmi 3.0 clocks?)
Reading up on optimal experimental design

Did some renaming and restructuring of the FMI2/3 VDM-SL models, with Kenneth.
Completed and pushed a bugfix for more sensible function value comparisons (that works better with lambdas)
Created first cut of POG methods in LSP Server for Jonas and Frederik
Added some POs to VDMJ that were missing from Overture (to do with inv/post satisfiability)
Helped Kenneth get annotations working (without using a plugin jar)
Fixed FMI-2-VDM’s toVDM() method to return a String rather than print to stdout
Look into an Overture bug of Leo’s that gets an incorrect warning for expressions like post_func[type].

Case study with Ken Pierce and Sam Hall on Digital Twin 3-joint robot arm.
Working on release 1 of Maestro2
Initiating proof of Mabl spec (termination)
Implementing CI and deployment for FMI2/3 VDM-SL models
Identified VDM2C issue - I am not continuing on this. What is the plan here? I think it is a pity to give up on this, as vdm2c creates MUCH better FMUs than overture toolwrapper and makes it possible to run it on microcontroller

Initiating proof of Mabl spec (termination) in UPPAAL with Casper
Studying the DSE scripts for genetic algorithm (sounds close the the third point of Claudio)
Finalizing the paper on formal analysis of drones at unipi

Overture
- Working towards Overture 3.0.2 Release
- A release candidate was produced and Nick tests found issues, that I will go over starting next week.
DT for water case
- I have been in a workshop organized by AVK https://www.avkventiler.dk/ as part of the the Danish Water Cluster initiave
- Next week I will participate in the kickoff of a platform for sharing data generated in the water sector: Innovationsdataplatform - Data Ecosystem Water workshop.
INTO-CPS Application
- Regarding INTO-CPS. We were able to upgrade the Electron version to the latest version! From 6 to 10.1.3! It required sorting out troubles with the download manager and dialogs, but we are now done with it. Kristoffer solved an issue Mirgita found and we are now working towards a new release and unit testing.
- Looking forward to collaborate with Maurizio in the improvement of DSE. Kristoffer and Hao can also join to both extend the app and create a new DSE project with the scripts from the paper about time-delay.
Other
- Working with Till on his survey on DT. Now and then he finds papers about digital twins in water, and I see good work comming out of the partnership in the long run.

Developed - together with Claudio - a dynamic model (mapping between motion and joint torque/actuator current) of an arbitrary robot with revolute (rotational) joints
The model is validated on the six axes Universal Robots UR5e robot arm
Working on expressing the dynamic model linearly in a set of dynamic coefficients for easy identification
When that’s done the rest of the work is easy and papers will be flowing

There was round table discussion on DIGITbrain experiments that use the INTO CPS toolchain.
No digital twin specific work in the last fortnight.