| Recap from meeting with Ariel | | Deb needs to meet with Sudhakar to go through the scope of the tasks, make necessary modifications, and figure out what the budget needs to be for the work that needs to be done. She is going to outline what we can deliver at the current cost and let Ariel make a determination if the cost should change.
ODE development is no longer needed. We will be modifying that to any V2X Hub changes that need to be made to support Tony’s projects or changes for OSS4ITS integration. We need to find the components to support reference implementation in the CARMA backlog and find a way to fund it to ensure it happens. We don’t need to think in terms of linking projects, we need to think in terms of finding where pieces currently exist, identifying gaps, and building them out while pruning off anything no longer relevant. We need to figure out how to make the most use with the packages we have.
Tony’s concept of architecture physical diagrams (as per their locations) was well received. Tony mentioned that there are some gaps. For e.g., big data implementation. We need to tie in that research being done with the others. It makes a lot of sense to have different implementation/code structure for different uses. For e.g., C++ makes sense on the edge solutions like V2X Hub. Java makes sense at the TMC level (JNI wrappers could be running on TMC/TIC and they are not developed using C++).
There needs to be a better low-level understanding of how the CARMA cars are going to be storing data. ICF project stops at a boundary what we can control. Rest of the functionality needs to be clearer on what is needed from CARMA and V2X Hub. We will set up a meeting with CARMA team to understand what is being done in CARMA and CARMA Cloud. Tony talked with Tim Tiernan about how we can pare down enormous files and get moving sooner. He was interested in taking an approach where V2X Hub might have a light duty computer engine for real time synchronized data feeds in JSON format. Deb thinks the idea of keeping a full data set for research has value, but not for every iteration of every piece of code all the time. We have to engage with the SDC team more. Currently the SDC functionality is built for federal users (using PIV card login) , and this needs to be changed for state users. They however use terraform scripts which have the capability to allow for regional groups or other DOTs to stand up an SDC of their own, but have internal active directory which cannot sync with state or federal AD. Lot of functionality is there that could be used for state/local implementation. We need to developed resources to be accessible by organizing areas that don’t include federal employees (i.e., don’t require a PIV card).
Deb would like to review the roadmap and see the critical paths defined where by elements outside the control of this project are identified and tracked, like SDC. She would like to pull out the developed architecture diagrams and use them for a separate project to show how it should be done. The key is to make sure that everything we need is in the current scope or is in changes what Deb/Sudhakar are going to work on. We also make sure that we are funded. Next steps for SET-IT diagrams – We have to wrap up CARMA updates, and also review any V2X Hub changes, after which Tom can review the overall diagrams again. Next steps would be to build Layer1 and then layer 0 diagrams.
Send out a reminder (and action items from this meeting) to SDC team (copy Ariel, Billy – Volpe) to remind them about the next Scum of Scrums meeting. |