A report from the International Data Corporation (IDC), a leading market research firm, revealed that the worldwide shipment volume of smartphones and tablets will reach 1.654 billion units by 2022. As the global demand for mobile devices exponentially rises, it is expected that location data will continue to explode. The challenge now is how organizations can harness the full potential of location data to enhance the customer journey and make better, more actionable business decisions.
This is where Mapbox comes in. As a leading location data platform for mobile and web applications, Mapbox understands the value of location data and how organizations can leverage it to generate new insights and unearth opportunities for business growth. Providing developer-friendly location data, SDKs, APIs, and beautifully crafted and lightning-fast maps, Mapbox helps businesses and their development teams gain a competitive edge.
SourceForge recently caught up with Ryan Baumann, a Solutions and Sales Engineering Leader at Mapbox, to discuss the benefits of an open source location data platform. Baumann also shares how Mapbox’s Vision SDK and Open Source Mapping Platform help developers and designers build a new world powered by location data.
Q: Can you give me a brief background of Mapbox (year founded, size, solutions, etc.)? What’s your company’s mission and vision?
Ryan Baumann, the Solutions and Sales Engineering Leader at Mapbox
A: Founded in 2010, Mapbox is the leading real-time location data, mapping, and design platform for developers, designers, and enterprises serving over 1,200,000 registered developers, and touching over 420,000,000 people monthly around the world. Mapbox has raised over $200 million in funding from leading investors including The Softbank Vision Fund, DFJ, DBL Partners, The Foundry Group, Thrive Capital, and Promus Ventures, the largest funding to date for any independent location platform.
Developers are the foundation of Mapbox’s success. With the largest developer community building on a live location platform, our mission is to empower developers to create the future of mobility by unlocking the full power and potential of location data.
Q: What does Mapbox specialize in? Who are some of your current clients?
A: Mapbox provides global maps, real-time traffic, location search, and navigation via APIs and SDKs. Our services power offerings from industry leaders across every sector of business, from consumer giants like Snapchat, Facebook, Tinder, and CNN, to enterprise titans like Microsoft, Tableau, and IBM. Mapbox is a foundation for other platforms, letting enterprises analyze their data, drone companies publish flyovers, friends find each other, and insurance companies track assets – among many other use cases and applications.
Mapbox is a major disruptor in geolocation, having upset the common notion of what maps are capable of. As we leverage location data to unleash human potential, Mapbox is building the live location data layer of the modern world. We’re creating a more realistic map of our world and its human activities — commerce, transportation, and connection with our environment and with each other. On a technical level, Mapbox is leading the future of where we know maps are going — beyond mobile, shedding interaction layers, and operating in a mixed reality paradigm where real-time, reality-grade location data power everything from gaming to automated vehicles to logistics.
Mapbox works with Bloomberg, Bosch, General Electric, Grubhub, IBM, Instacart, Mastercard, Metromile, Microstrategy, Square, The Weather Channel, Uber, and 600 more enterprise customers who rely on our products and tools to shape the way 420 million people – their customers – explore the world.
Q: Tell us more about Mapbox as an open source mapping platform. How does Mapbox support every aspect of the web and mobile map-making process?
A: Mapbox proudly embraces our open source roots. We work in the open and release as much code as possible. We currently have more than 660 public repositories on GitHub, most licensed under BSD. We do this because we believe that it’s the right thing for people, technology, and business.
By releasing our software as open source, we’re hoping to encourage the use of reusable, general-purpose parts. In many cases, this means modules rather than applications. We distribute these modules for other developers to use with package managers like NPM and PyPI.
Q: In your opinion, what makes for a reliable location data platform?
A: We think of “reliable” platform as a service at scale that can hit millions of users, works anywhere in the world with the same SLA, and never has downtime. We embrace open source, and platform is the core of what we do. (mapbox.com/status)
Organizations that put trust in products powered by Mapbox APIs can build their business on our platform. We help companies by instituting for scale on top of AWS services. We host our platform in every AWS region – across five continents, 11 countries including China, and 9 isolated regions- by distributing data through 60 edge cache locations. (mapbox.com/platform)
We believe in transparency and automation to take the human out of the loop whenever possible. If there is ever a problem with service, we will know about it before you do. (status.mapbox.com/history)
Q: What makes Mapbox stand out from other mapping platforms available in the market?
A: Mapbox stands out from other mapping platforms such as Bing, HERE, TomTom, Google Maps API, and to some extent, Apple Mapkit because we don’t force users and enterprises to use off-the-shelf maps.
Mapbox is a flexible, open source platform that has become incredibly popular among developers because we allow them to create custom mobile and online maps that fit perfectly into the environment they are building. Here’s how we’re different:
- Platform layers. Mapbox is different from other mapping and location service providers as we don’t just offer a static map, we offer a platform organized in over 130 data layers which allow the developer to fully customize their maps by picking the layers they want for their application, and/or layering in their own data.
- GL game technology. Mapbox uses video game technology to design and render the map at 60 frames per second – which no one else in the industry does. This allows the maps to be animated with fly-throughs, easy zoom, scroll, and other experiences that look like live action video or game quality to the end user.
- Customizable design. Mapbox is well-known for its amazing design tools that enable developers to create beautiful, customized maps. We allow customization of what shows up in the map (data layers), how it shows up (colors, shapes, icons, fonts), and we have a rich library of styles that designers can choose from for speed and inspiration.
- Richness of data. Mapbox is unique for the richness and ever-improving nature of our data platform. We collect over 250 million miles of anonymized road data every single day and that number is growing. Further, we enrich our platform with over a hundred different open and proprietary data sources, including satellite imagery that we process and enhance, community activities that create new open source map data in remote areas, and other data from around the globe.
Q: As advocates of the open source framework, why is an open source approach beneficial for designers and developers?
A: Switching from the proprietary APIs to an open source approach gives designers and developers the ability to choose between new, open, and less expensive mapping alternatives like Mapbox.
In a nutshell – it provides control over the user experience. Having the power to obsess over every detail about your product is invaluable.
We believe in breaking down barriers to building. Open source gives our developers the power to integrate with all the other tools in your developer toolbox without any limitations. Throughout our growth, we have stayed true to our developer-first, open platform roots. We obsess over extensible, flexible tools that give developers full control of their location-based experiences.
Q: Open data is a core part of Mapbox’s platform. How do businesses benefit from utilizing an open source location data platform?
For example, IBM incorporates the open source mapbox-gl library into their Java Rave rendering engine that power Watson Analytics.
Companies like DoorDash can integrate navigation and ETAs directly into their mobile `Dasher` application, so that you can check exactly when your food is going to arrive.
Also, MicroStrategy uses Mapbox Enterprise Boundaries and Mapbox Streets, built on top of open data sources like OSM, to help their business intelligence users understand their location data in dashboards.
Q: Can you give us some examples of your open source tools? What are the strengths and specialties of each?
A:
- Mapbox GL is a suite of open source libraries for embedding highly customizable and responsive client-side maps in Web, mobile, and desktop applications. Maps render at a super high frame rate. You can use custom styles designed in Mapbox Studio to manipulate every aspect of the style’s appearance on the fly – like changing the color of a category of restaurant based on a user’s food preferences.
- Turf.js is a geospatial toolbox for javascript web developers. It brings most of the power of PostGIS spatial functions to any web or Node.js application. This allows developers to build user experiences with GIS calculations like point-in-polygon and buffer anywhere in the user experience, not just in the database.
- Rasterio wraps the powerful features of the Geospatial Data Abstraction Library (GDAL) in idiomatic Python functions and classes. If your organization is working with satellite or drone imagery data for construction, agriculture, or disaster relief use cases, you’re probably using Rasterio. It is the most active Python project in Mapbox’s GitHub organization and one of the most active open source Python GIS projects.
Q: How can developers manage the risks associated with utilizing open source components in their applications? Is there any particular vulnerability developers must be aware of?
A: Open source projects are only as strong and secure as the people that maintain and support them. Mapbox significantly invests in its hundreds of open source repositories. Internally, we have an open source standards team that reviews every new project we choose to open source.
Software needs to evolve to work with other technology and security standards. If an open source project doesn’t have an active maintainer, there is no direct ownership or protection from a new security flaw. As a business, if you need to fork an open source project to maintain it, and that is a risk your business chooses to take on.
Using OSS with an active security vulnerability disclosure policy or bug bounty program is one way for developers to manage OSS risk. At Mapbox, we maintain a security vulnerability disclosure policy for our software, including our OSS projects. We publish security bulletins for our open source projects on our website’s security bulletins. We also submit reports of security issues to npmjs. We maintain a bug bounty program on HackerOne and our open source projects are within the scope of this program.
Q: At your recent user conference Locate, Mapbox announced the release of your new computer vision product, the Vision SDK. Can you tell us more about this solution? What are its key benefits and capabilities?
Mapbox Vision SDK
A: The Vision SDK turns a smartphone camera into a new set of “eyes” for your car. It enables utilization of imagery from connected cameras on “the edge,” bringing visual context to Mapbox’s live location platform and empowering developers to build exciting new applications.
Mapbox Vision SDK enables a suite of features: it powers augmented reality navigation with turn-by-turn directions, safety alerts (for pedestrians, vehicles, and cyclists), sign detection (speed limit, construction, school zones, etc.), and live updates of how the driving environment changes.
On the back end, the Vision SDK provides developers with invaluable metadata that feeds into the living map. Real-time road conditions and event data are automatically geo-coded and can be used to improve map fidelity at global scale, without needing to wait for imagery uploads. Neural networks power feature extraction that runs on the edge, so bandwidth needs are minimal.
As for what it can actually “see,” the Vision SDK does three different kinds of computer vision tasks:
- Classification. Looks for the existence of something in an image frame (like a specific traffic sign) and indicates whether it is there or not.
- Detection. Looks for discrete features within an image frame (pedestrians, for example) and identifies each instance with its location.
- Segmentation. Looks at every single pixel in the frame and assigns it to a different class (the road, lanes, curbs, etc.)
The Vision SDK can be used for cars, vans, buses, etc. Other IoT devices, such as drones, traffic cameras, robots, and others will be supported. Motorbikes/scooters are also on the roadmap.
Q: Aside from Vision SDK, what can we expect from you in the future? Are you brewing up any new product or any new product or offering?
A: We’re focused on expanding our market efforts around our existing products and also to further commercialize technologies we have already created. In the coming year, you will see Mapbox further expanding internationally, in the automotive space, in business intelligence, logistics, AR, data visualization and other key areas. Further, we will continue to invest in improving our data sets with POI, address, imagery, and automotive data that will enrich our global platform and enable new, previously unimagined services.
About Mapbox
Mapbox is the leading location data platform that provides global maps, location search, real-time traffic, and navigation via APIs and SDKs. With a mission to “change how people move around cities and explore the world”, Mapbox focuses on designing, building, and developing tools and applications that enhance user experiences and make lives better. The company specializes in on-demand logistics, asset tracking, store locator, turn-by-turn navigation, and data visualization.