Home Blogs Geospatial Industry Trends 2017: Here’s what leaders have to say

Geospatial Industry Trends 2017: Here’s what leaders have to say

Geospatial business leaders discuss the future of geospatial industry at Geospatial World Forum 2017

Geospatial Industry Trends 2017
L-R: Andreas Gerster, Alain De Taeye, Walter Scott, Steven W. Berglund, Mladen Stojic and moderator Karthik Ramamurthy

Your newsfeed these days is bound to be filled with words like artificial intelligence, digital engineering, augmented reality, business intelligence, etc. And while the common man grapples to understand these technologies, stakeholders of the geospatial industry are being challenged by rapid developments in these fields.

With the infinite computing capacity of the Cloud, automation has been regarded as the answer to derive intelligence from today’s explosion of data. From one perspective, it even looks like the machines are taking over our industry. The time seems right for the geospatial industry to acknowledge the rise of smarter machines and embed computing into the workflows for added insight.

At Geospatial World Forum 2017 in Hyderabad, industry leaders discussed how leveraging artificial intelligence not only provides greater understanding of the complex data; it opens up plethora of opportunities and challenges alike to the industry. Read what they had to say…

Alain De Taeye

Founder, Tele Atlas, and Board Member, TomTom

On IoT and map content

We talk about the Internet of Things (IoT), but Internet of Things doesn’t make any sense unless you know where those things are. Maps are crucial in this. The whole society today depends on maps. Three decades ago, navigation was considered science fiction. Now, it is a commodity of everyday use. And maps without mobility are unthinkable. In fact, new mobility services are fast becoming fully operational. I even believe that self-driving cars will happen and they will happen much sooner than people believed in the old days.

In maps, the number of layers is increasing, the content is increasing. But one critical aspect is posing a challenge — we don’t have information up-to-date to the minute. That’s the biggest challenge ever. And the only way forward is to automate processes by which we make and update maps.

Traffic Sign Classifier AI

Now, we all know traffic signs are essential for maps. But mapping them earlier used to be a long process. TomTom has automated this complex visual process using lasers. Our Traffic Sign Classifier AI has created a traffic sign evidence database of 100 million signs. And this is just one of the thousands of applications people will be using in the future.

I suspect that the maps of the future will be fuelled by communities, sensor data, professional mapmaking services, automation, and deep learning. AI-based automation is critical to enable high efficiency and short cycle times. Moreover, scalability — which is needed to deal with the global demand and increasing content — is possible only through AI. So, a transactional mapmaking platform is imperative.

Steven W. Berglund

CEO, Trimble

Intelligent machines and AI

An intelligent machine is a flexible rational agent that perceives its environment and takes actions that maximize its change of success at some goal. So, this implies understanding both, the context and the desired outcome. The number of connected devices is growing at a such a massive rate that tere are more devices today than there are people. The market opportunity presented is huge. All markets are being shaped by intelligent systems in a more connected world.

IoT and Cloud have created the perfect storm for general purpose AI. This is because AI unlocks the potential of sensor data generated by IoT. By 2020, a smartphone is likely to be more smart than today’s super computer. Increasingly, with the computing power and systems capability, you talk about improving work processes. Ultimately, you start to look at things holistically — in terms of enterprise productivity.

Precision farming and transportation

Sensors and AI accelerate precision farming by monitoring crop conditions, such as, water, nutrient, plant population, soil moisture content, etc. Disease prediction becomes easier. Automated variable rate irrigation becomes possible, along with autonomous tilling and harvesting.

In transportation, computer vision and Big Data can be used to predict road conditions and transportation demand. Sensor integrated telematics systems help in optimal fleet deployment. Machine learning and data intelligence enables preventive maintenance. So, integration of smart infrastructure with smart vehicles provides significant opportunities.

Information process and its implications

There have been several changes in the information process. Things that were done in the office are now done in the field. Crowdsourcing has become a viable source of data for many applications. Access of data is moving from proprietary to being free and ubiquitous. Applications are becoming embedded in the process. Update frequency has changed from version-controlled to dynamic.

This has led to several considerations and implications we need to take stock of. These include organization design, such as, span of control, accountabilities, etc. Monetization and the potential diminishing returns of information also need to be considered. Another important factor is the consequence of mission critical failures, including system complexity, security and human/system handoffs.

Mladen Stojic

President, Hexagon Geospatial

Beyond traditional models

We live in a dynamic world that is multi-dimensional and quite complex but the technologies we have today result in a static representation of that world. Current tech focuses on establishing sensors and software that allows us to do pleasing imagery. But that is still a static representation, and we have to move beyond this static understanding. Time is a critical factor. Once you have that 4D element, you can move beyond the traditional models.

Influencing positive outcomes

We need to not only understand what’s happening and what has happened, we need to know how to influence an action. AI and deep learning hold the power to influence a positive outcome. One area where we are seeing a lot of growth in is use of robots and UAVs – for the purpose of delivering more information to support a critical mission at hand.

We spent many decades understanding what was, now moving to an age understanding what is, now entering a new generation of understanding what can be – that’s where the real role for deep learning and advanced analytics lie.

Walter Scott

Founder and CTO, DigitalGlobe

Data management challenge

Remote sensing began with a quest of providing global transparency. Today, remote sensing is becoming ubiquitous and it is creating a lot of data. Which brings us to the data management challenge: DigitalGlobe has amassed 100 petabytes of data since 1999. This kind of data would crush any IT dept on the planet, and our ability to extract useful information at scale from that data. The technologies helping us in this are elastic Cloud Computing, deep learning and crowdsourcing.

Authenticating crowdsourced data

When I was a grad student, everybody hated statistics. Nobody wanted to pick up that subject. Today, statisticians are rockstars. What we use in crowdsourcing are some advanced statistical techniques that effectively let us know when people are accurate and when they are not. It’s a matter of letting large amounts of data help sort us through the noise. When you create a feedback link between the crowd and the machine, the machine acts as an element of the crowd.

Andreas Gerster

Vice President, BIM/CIM and Product Design, FARO

Driving efficiency

It all starts with capturing the reality. Capturing the real world is one thing, but what do you with all that data? You need to get the information out. So long as the data is only available on one computer, it is of limited value.

Now, technology is a great tool, but it’s good to have a fallback on a human being. I want to stress that FARO is a modern IT company. We are observing the tools, but we are not driving the technology. We want to make people on the construction side more effective and efficient. We want to enhance the human brain by providing info that is available.

Data privacy

It is not just a privacy issue, but also a security concern. Imagine you have reality scanned data from an atomic plant, and it gets in wrong hands, what will happen then? We have to think about this. It also depends on the region. Customers in US are much more willing to share data than customers in Europe. So, we provide different platforms depending on the needs.