Using Maps to Make Autonomy Work for All

TL; DR

  • Autonomous vehicles must work for everyone. Not just tech companies. Not just the auto industry. Not just those that will be able to afford it. If it’s going to work, we need to work with everyone this technology will touch: from passengers and pedestrians, to governments and communities.
  • The keys to doing that: a People-First approach to safety and accessibility, proactive Data Transparency and concrete commitments to Economic Development.
  • Maps that autonomous vehicles depend on have turned out to be an effective way to serve all three of these tenets without stifling progress.

Interested in joining the front lines of AV mapping? Passionate about redefining transportation and cities? CARMERA is hiring.

Photo credit: Bosch

Autonomous vehicle (AV) platform, Waymo, has self-driven more than five million miles and recently announced they will open their fully automated service to the public this year, beginning in Arizona. Earlier, self-driving taxi company, Voyage, began its work to bring services to the country’s largest retirement city, located in Florida. And Governor Andrew Cuomo recently extended statewide legislation through 2019 for testing, right here in New York.

Tangible progress continues in the AV space, and is happening at an increasingly fast rate. Overall, that’s positive. But as the recent accident in Tempe, Arizona in which a pedestrian was struck and killed by a self-driving vehicle showed, there is much work to do.

Part of figuring out what that path looks like is defining what we’re here for — and that starts with a universal mission that if autonomy is going to work, it must work for all. It must work for cities and governments as much it will work for the auto industry. It must work for underserved communities as much as it will work for the world’s tech hubs.

We’ve recognized that in order for the industry to deploy this technology at scale, a more holistic way of thinking is a requirement, not a choice. This isn’t about pushing some code and “it just works.” AVs create so many externalities, most of which are very promising. But since those externalities affect everyone, they need to be accounted for upfront.

So how do we do it?

The short answer is collaboratively and practically. From engaging directly and frequently with all of the above stakeholders, both in the U.S. and abroad, we at CARMERA firmly believe autonomy works for all when our industry becomes committed to three core tenets: People, Transparency, Economics.


Design Around People First, Always

Beyond auto and tech circles, many see autonomy as at odds with humans. People have (valid) concerns when it comes to safety. They have questions about affordability and access. But in order for autonomy to work, it can’t be pitted against people. Autonomy has to be “people first.”

If you stop and think about it, autonomous mobility is actually entirely about people. It’s about equitability. It’s about passenger and pedestrian safety. It’s about bringing affordable transportation to underserved populations that need it most; populations that don’t have reliable access to transportation to get them to their jobs, school, services and everywhere in between.

So if AV is going to embrace this, every aspect of the technology that powers autonomy needs to be designed around people-first principles from the outset, as opposed to an afterthought.

We have found that AV mapping systems can play an important role in making good on the people-first promise. AV maps that are accurate and fresh make city streets safer for all. For passengers, they are considered a requirement to achieve the level of redundancy and reliability AVs will require. It’s a bit like laying down and pressure-testing rail tracks before a train is allowed to run on them. For non-passengers (those walking, cycling, scooting, hovering, etc.), dynamic map data can play a critical role in directing AVs to avoid the riskiest areas. Take, for example, pedestrian density data. Through our network of delivery fleets outfitted with sensors, we are able to gather statistically accurate observations of which specific blocks in a city pedestrian presence tends to be heaviest at different times of the year, week or day. Our AV customers can then take this important data into account in routing along with traditional inputs like distance and speed limit.

A key part of people-first is doing all of the above responsibly, by protecting sensitive information and adopting clear policies upfront that prioritize privacy.


Embrace Win-Win-Win Data Sharing

Every advanced, large-scale transportation technology — from shipping to railroads to aviation — has converged upon an equilibrium for sharing critical data. It will be no different for AV. We must prove that the safety-critical data powering our tech is verified and safeguarded and that the “data exhaust” coming out of our services can be made useful to all (i.e., city planners, property managers).

This isn’t a call for AV companies to reveal trade secrets or sacrifice competitiveness. It’s a call to proactively respect the institutions we must engage with by sharing innocuous, yet high value data. We recently realized that maps can act as a natural information nexus to this end.

Humans have been reading maps for centuries. We trust maps. They are designed to represent the world as accurately as possible, especially the high fidelity, real-time 3D maps that AVs require to confirm where they are, what they’re seeing and where to go next. So when it comes to sharing information that is both critical for safety verification and useful for everyday planning and analysis, data that goes into AV maps (such as 3D spatial models and pedestrian density statistics of streets, per above) are perfect ways to deliver on this core transparency tenet without commercial risk. We are playing our part in this at CARMERA by making our data accessible to governing agencies of all types, including an upcoming collaboration with local Brooklyn institutions, and we will be doing more internationally as well.

3D scene reconstruction data that can be used for urban design and construction, in addition to AV maps

Jobs and Economic Impact: Have a Plan

San Francisco Chronicle tech reporter, Carolyn Said, penned a thoughtful piece last year addressing the impact autonomous vehicles will have on jobs, noting that for “millions of people that drive for a living, autonomous technology could make their work obsolete.” Is this true? Potentially, yes. But the article goes on to explore what job creation potential AV holds. By working collaboratively with local communities, AV companies are poised to create new high and middle skill jobs that are better wage and lower stress. But what’s missing for many is an actual plan. No one is saying the plan can or will be perfectly prescient, but lawmakers and workers need to at least know how to connect the dots in order to set their communities on track to reap the potential benefits of what could indeed be a new industrial revolution.

When it comes to our role as a mapping company, we’ve made concrete commitments to economic development programs like START-UP NY and Urbantech NYC. These programs represent our commitment to growing local jobs and generating local capital investment. We’re now doing this out of our Seattle office too. Some of those jobs include roles in robotics, machine learning, computer vision, geospatial, data science and other engineering disciplines. Others include operational needs in sensor maintenance, GIS analysis, data annotation, field ops and more. We are also actively engaged with academic institutions like NYU, University of Washington, University of Michigan and Michigan State University to make sure we are investing in the local talent pipelines where we operate, through internship and data sharing programs for educational benefit. We are doing all of this and looking to do more in areas like retraining because we believe in it and because it’s good for business.


In Closing

For autonomy to work, it has to work for all.

We are committed to ensuring that maps play a significant role in getting us there, and would like to hear from you if this sounds relevant to your own community or business, or if you have suggestions for what more we should consider in the future.