Geo behind every corner, even if you don’t expect it. Predictions by Ernest Earon, CTO of PrecisionHawk
Is “GIS is everywhere” a redundant statement? Aside from the rather poor (though literal) pun, I argue that it is redundant. I propose that GIS is currently going through a renaissance period of explosive growth, though perhaps not in the most obvious of ways.
To the greater public, at least those who are aware of what ‘GIS’ is, the field often seems to be highly specialized, complex, and niche. It is the domain of highly skilled technical people using sophisticated tools to track, catalog, and (ideally) understand the world as it unfolds in space and time. While that is definitely a big piece of what is happening, we are also in a transformative period where ‘GIS’ is being incorporated in a myriad of often subtle ways that are changing both what and how we know. What is happening is that applications and use cases for spatial and temporal information and analysis are appearing in great numbers at a faster rate than ever before. As such, core techniques and technologies are being isolated and applied everywhere.
This is being driven by several factors:
1. Consumer Demand
This is likely both the biggest, and the most subtle, driver for the widespread use of GIS technology. It’s also where the general public might look and say “That’s GIS? Cool.”
This is very much akin to how computational vision processing went from being the rarified, specialized domain of PhD’s in university labs (and a tremendous mouthful to start a conversation at a party with), to something that is virtually taken for granted now. Look at how your camera or smart phone can automatically detect faces in a scene, maximize the number of faces in that field of focus, and even trigger the shutter when people smile. The same holds true for how those devices attach metadata to those images. Cameras (whether in a phone or stand-alone) now commonly contain a GPS device to tag photos so that they can be more readily cataloged and queried later. While you might say that that’s hardly high-grade GIS technology, it is applying spatial and temporal information to other data to improve the end use or workflow. This might be a somewhat trivial example, but is hardly the only case where consumer applications is driving this technology.
Aerial imagery and drones are another great example. GIS for processing, analyzing and storing this information are a critical requirement. While a live-streamed video of a snowboarder going down a half-pipe might not immediately strike one as needing this technology, being able to effectively replicate the route or reconstruct a 3D model of the course does.
Certainly when using drones to collect aerial imagery GIS techniques are paramount. Anyone who has looked at 750 sequential aerial images of a corn field knows that that is not a useful format for the data. It is not just critical to be able to properly create a rectified, accurately referenced mosaic, but something needs to be done to process and extra useful information from that data set. Anything from hydrology models to timber inventory estimates.
2. Technology (collecting and computing)
Drones are not just a key piece of the puzzle because they enable a new data collection method. They are also a great indicator of a larger trend: that of rapidly improving technology at an increasingly reduced price. This not only puts aerial imagery capabilities in the hands of anyone from pre-high school kids all the way through to large existing manned aerial servicing outfits; it also is increasing the number of sensor modalities those people have at their disposal. 5 or 6 years ago Tetracam was pretty much it for low cost multispectral imaging. Now there are easily a dozen options on the market, and more are coming. Alongside this, the performance and cost keep improving. This driving even more applications and more data collection to be undertaken.
Computer resources are also continuing to improve. The Internet of Things movement has seen tremendous growth in embedded controllers and signal processing capabilities. This allows processing solutions to be embedded, speeding up the time-to-use of the data, and enabling more and innovative use cases.
Indeed, the commoditization of all of the technologies involved is opening a great many more opportunities and applications.
3. Commercial and Industrial Applications
Another key feature driving the growth of GIS, is the expansion of commercial and industrial applications for the data (for the record, this is hardly an exhaustive list of the thing making GIS technology a must-have for the remainder of the 21st century and beyond. It’s really just a few examples).
The examples above have really enabled and demonstrated the disruptive power of knowing more about the right place at the right time, and companies are rapidly focusing efforts to harness that to improve their business. Drones are a great example here, too. While 10 years ago using a small robotic aircraft to do routine automated storm damage assessments would have seemed ludicrous, it is not accepted as critical to best-practices going forward. Companies large and small are rapidly standing up and heavily investing in groups to make this a part of business as usual.
Making remotely sensed data a standard part of your operating model is simply accepted.
Of course, right alongside this is a huge need to build the techniques and implement the tools to make use of that data. Algorithm development is now moving at a pace we haven’t seen since satellites came on the scene. In fact, it’s unprecedented and is dwarfing the to date. Now that everyone can collect the data, and everyone needs it, it’s clear that GIS and remote sensing is in a bit of a golden age. Albeit a tremendously busy one.
The other feature that is consistent through all of this, and what is making this truly a standout period, is that all of issues are reinforcing each other. People are simply expecting better performance and results that require GIS analysis; the tools to collect more data that needs to be analysed are cheaper than ever; the applications for the data are exploding; and businesses are demanding more of it.
While it is difficult to predict what this will look like in 10 years, it is clear that the widespread solutions will not be driven by analysts in front of workstations. Instead, it will be tools and techniques developed by those analysts being embedded in everyday technology seamlessly integrated in our daily lives.
That’ll be handy when your self-driving Uber is monitoring tree growth to predict pollen intensity and adding that to the antihistamine global logistics chain. And reminding you that the pharmacy at the next corner still has some…
I’m thinking of going on a theme of “GIS is everywhere.” It’s not niche at all. Much like image processing technology used to be niche, now it’s on every camera, every phone, Facebook, Instagram, etc. finding faces (to focus the camera on) and identifying people. It’s a capability that we now take for granted. GIS is like that. Then, with companies like PH driving applications and lower cost adoption, it’s going to get bigger. I see the idea of a dominant GIS player giving way to smaller actors embedding the tech and capability into everything.
Designing “Irresistible Web APIs” – book review
More and more businesses are creating web APIs. This is because APIs can add business value to a company as it offers developers an opportunity to interact with open platforms of others, for example by embedding external sources as social media on a website. The geospatial industry is has also embraced APIs: some companies are built around APIs (Mapbox, Carto), while others are now adopting an new approach to APIs, such as Esri who recently released a ‘pythonic’ API that allows for interaction with ArcGIS Portal and ArcGIS Online. Lately, there has been an increase in mapping APIs, as a result of Google ending its support a component of their mapping API called Google Maps Engine.
The growth of web APIs is self-reinforcing: as more web APIs become available, more people realize the potential of having their own API, which leads to more APIs being created. But not all APIs are instant successes, and even successful APIs may one day be deprecated, as the example of Google Maps has shown. The question is, then: what makes a good web API?
A recommendable source that answers this question is Kirsten Hunter´s book “Irresistible APIs: Designing web APIs that developers will love” from Manning Press, published recently. After having been involved in the API design process and implementation of large companies such as Netflix and Twitter, the author realized that these use cases might inspire others in designing web APIs.
The book is divided in two parts: the first part is called “Understanding APIs” and covers technologies and best practices for API creation and design, while the second part features an overall strategy for API design and creation. She starts off with a discussion on the overall goals and ideals for an excellent API program and offers a high-level description of technologies and techniques used for web APIs, as well as best practices for excellent APIs. Readers are encouraged to actively interact with a simple web API that allows to add and create pizza toppings on a virtual pizza, monitor HTTP browser traffic using HTTP sniffers and make API calls through a software program named curl.
API design principles
The second part of the book, titled “Designing web APIs” is the center part of the book. Important API design guiding principles such as “don´t surprise your users”, “focus on use cases” and “focus on the developer experience, not the architecture” are exemplified with real-world examples. The value of an API and how you measure this, combined with creating meaningful use cases is another important part of the API design process. For company´s whose only products are APIs, monetization is an essential business goal but in other cases, such as where an API is not a main product, there may be other business goals such as usage or partner retention.
Once you know why you need an API and what it needs to do, you can start modeling the schema of an API. The basics of schema modeling are explained through the two most popular schema modeling systems, RESTFul API Modeling Language (RAML) and OpenAPI format (previously named Swagger). API Development and project management strategies are covered in a chapter on design-driven development and help you building your API. A final chapter on providing a developer experience that makes your API stand out. This is a crucial part in API design and the success of your API depends on it.
This book is best characterized by a quote on the back cover, stating that “a future is emerging where our APIs are our public face as much as our website, office space or sales staff. This book points the way”. What the author of this book has noticed correctly , is a change in software architecture design over time where an API is not just an add-on, but an important part of that same architecture that serves to extend a company´s main product and can generate more revenue or users. This change also requires a more integrated design approach when designing and building an API. The advantage of successful use cases allow others to build on, as well as open APIs from company such as Facebook. The design principles and use cases are the main attraction of this book and the reason why I´d recommend it.
The book is very accessible, judging from the non-technical language and examples. This makes it suitable for a wide audience, but the downside of this approach is that software developers hoping to find a technical book on API development will probably be disappointed with this book. This is because it is not a tech book per se, but one that offers a high-level overview of how you design APIs before you build them and all organizational challenges that come with API design. There is little technical content – it’s not worth reading the book for technical content alone.
Unfortunately, the first part of the book is not as good as the second part. The reason for this is that the book lacks unambiguous definitions for key concepts as ‘APIs’, ‘web APIs’ and ‘REST APIs’. The author explains the term web API but uses different interpretations in different contexts, which is not helpful in understanding the concept of web APIs. I would have preferred a more in-depth and technical approach during the first part of the book. Here, the author seemed to be in a hurry in order to quickly serve the main dish, which is the part that covers API design. Even though this is not a perfect book, it´s recommend for the chapters API design and use cases.