Could MDS expand scope to include collisions and complaints?

[mateoclarke]

I brought up this question in the Webinar hosted by LADOT today and looked for reference in other issues but didn't see anything yet.

City of Austin has adopted new rules as of November 9, 2018 that requires a real-time and historic API according to our Director's authorized specification (which is currently MDS). It also requires monthly reporting for complaint history and collision history.

I'm wondering if other cities are including rules like this (collisions and complaints) and if there is an appetite to accommodate this type of reporting into the MDS in future versions.

[hunterowens]

Thanks @mateoclarke! I think one of challenges in managing

re: complaints, I wonder if that should be an Open 311 datatype, rather than MDS. LA at least is collecting complaints via 311. I know the CAN letter recommends using 311 to do service requests but if city are expecting the vendor to track complaints it would make sense to add into provider.

re: collisions, my worry here is that is very hard to find/track collisions because they have to self reported and therefore hard to compare data from across operators. If there are features inside the hardware itself that would be able to report back, I'd be interested in adding it.

[asadowns]

I would encourage MDS to become the source for complaints and collisions. I think MDS spec is a good home for complaint and collision endpoints with a couple caveats.

• Concerns around data retention and PII are addressed as these are sensitive in these cases
• The data reporting requirements should not be onerous
• Providers should not be punished for being proactive about collecting data, volume of complaints is not an acceptable metric for evaluating providers. For instance providers that provide a friendler, more prominent complaint submission experience will receive more complaints.

I think a good format for this could be similar to status_changes with providers notifying agencies when complaints are open, in_review, reviewed, and closed. With a minimal amount of metadata including location, unique id, complaint type (damage, parking, riding), device_id (when available) and an optional area of reasons.

Incidents would be handled similarly but with the type of collision, injury boolean, and what was struck (person, vehicle, stationary object).

[black-tea]

Hi @mateoclarke - Regarding collisions (not complaints), below is the current view of LADOT. If cities find that the following summary doesn't adequately represent their situation, we encourage them to begin thinking about additional standards, but we will not be actively prioritizing the development of MDS-based collision standards.

re: complaints, as mentioned by @hunterowens, we've been thinking about this but haven't yet settled on anything definitively. Thank you @asadowns for the suggestions of what we should be thinking about; we are happy to keep that conversation going.

---

Background The Los Angeles Department of Transportation (LADOT) relies on collision data collected by law enforcement agencies and released as part of the Statewide Integrated Traffic Records System (SWTIRS). SWITRS is a product of decades of testing and is informed by the Model Minimum Uniform Crash Criteria (MMUCC), which takes input from NHTSA, FHWA, FMCSA, NTSB, the GHSA, and subject matter experts from State DOTs, local law enforcement, emergency medical services, safety organizations, industry partners, and academia.

Shortcomings of MDS for Dockless Collisions Collision Data Standards Already Exists: MDS was created to address a specific gap in data specifications (trip & operational data for dockless devices), which is not the case for collision data. All states already have collision standards and record management systems in place, similar to SWITRS. Given a few minor adjustments, standard collision reporting methods can easily incorporate collisions involving dockless devices.

Privacy Concerns: Collision data always includes sensitive personal information on all parties involved in a collision for better understanding of contributing demographic factors. MDS, in contrast, is specifically focused on device information, not users. Any MDS-like implementation for collisions would be able to capture only a fraction of the detailed data currently collected by law enforcement agencies without raising additional concerns about user privacy.

Uneven reporting by provider: Providers that are more proactive at collecting data from their users will likely report a higher number of collisions compared to companies that make no effort to collect data or, more perversely, attempt to obfuscate such data in fear of more stringent regulation. Although the current collision reporting process likely undercounts the number of collisions involving people walking and bicycling, reporting by provider would add additional bias to the counts.

No capacity to verify or audit data: LADOT does not have any capability to verify the integrity of the data reported by dockless mobility providers. Device locations and statuses may be ground-truth verified by going out and inspecting devices. However, it would be nearly impossible for public agencies to verify the accuracy of event reporting by mobility providers.

Long term sustainability: Transportation departments rely on longitudinal collision data to understand long-term trends. Collision data collected by MDS APIs risks being lost if dockless mobility providers are no longer in business and cities had not stored collision information from the provider API.

Plan: Amend CHP 555 to Improve Data Collection Given the difficulty in developing an additional set of specifications to cover collision data, it will be much easier to amend current collision data standards to include collision data related to dockless devices. LADOT has already included proposed reforms to CHP Form 555, the standard form used for collecting traffic collision information in California, to include electric scooters as an additional vehicle type.

[hunterowens]

closing this issue for now.