exactEarth has successfully launched an earth observation satellite constellation to track Automated Identification System (AIS) data from the world’s shipping traffic. There are a variety of applications for this technology, including safety, supply chain tracking, and military operations, as well as environmental monitoring. Sensors & Systems (S&S) special correspondent Matteo Luccio spoke with Graham Stickler, senior director of Global Marketing at exactEarth, about the different markets for this data and services, the interesting applications, and the implications of this global insight.
S&S: What’s your background?
Stickler: I’ve been in the geospatial industry for 30 years, in various roles, in both the public sector and the private sector. My career started in satellite remote sensing in Africa. Then I moved into GIS when it first became popular in the early 1980s, as a means of disseminating and using remotely sensed data. I spent time in the water industry and working for GIS vendors—I worked for Esri, Unisys, and Laser-Scan/1Spatial. I’ve also done a lot of work on the geospatial database side at a technical level, with ArcSDE, Oracle Spatial and SqlServer Spatial. I’ve worked a lot with open standards, standards adoption, geospatial data quality, and geospatial database management and I have been involved as a working group Chair at the OGC [Open Geospatial Consortium]. I have a wide and varied industry perspective; the longest period I spent in any one specific market is in transportation and asset management.
S&S: Do you have any nautical experience?
Stickler: I don’t really have too much previous nautical or maritime-specific experience, although I do have extensive experience in transportation, asset management, charting and mapping, and geointelligence, all of which I’ve found are very relevant. I was brought into exactEarth primarily because of my geospatial systems and data background as while our IP is focused around satellite systems and data collection, we actually deliver geospatial data to our customers. We are a young company and to begin with we focused our energies on creating the infrastructure, including launching a constellation of satellites, to collect and process AIS data from the world’s shipping. We’ve moved forward and now we are very much looking at how we supply our customers with these data, which is very much a geospatial data challenge, as opposed to a satellite challenge. I was one of the first people recruited into exactEarth with geospatial background to help the company move forward in that manner.
S&S: What does exactEarth do?
Stickler: We use satellites to collect the messages that ships transmit via AIS, which is a technology under the SOLAS [Safety of Life at Sea] convention of the International Maritime Organization—the United Nations specialized agency with responsibility for the safety and security of shipping and the prevention of marine pollution by ships. Every ship in the world over 300 gross tons or that carries fare-paying passengers must carry an AIS transponder and these devices transmit to each other for the primary purpose of collision avoidance. Ports in particular use AIS as a maritime traffic management system and have set up terrestrial receivers to collect the data in and around ports. Being a radio frequency-based system, these devices have a limited range, primarily due to the curvature of the Earth so away from port. Ships still literally disappear over the horizon. exactEarth was formed on the basis of the belief that the AIS signals could be collected from space, which has no horizon. You can see everything without limits from above and we, as a result, pick up a global picture of the world’s shipping.
S&S: How do you present this technology to potential clients or users?
Stickler: AIS data are made up of a collection of standard defined messages that ships transmit within any one time slot about who they are and their position, speed, course, etc., many of which are entirely auto-generated by the ship’s on-board systems. There is then also a host of other information transmitted as binary messages defined by the IMO, and there are regional binary messages defined for specific administrations. These binary messages are essentially free-form, so the content is “as required”.
We tend to find that we have two classes of users. For many, details contained in the standard messages are enough and allow you to track the ships. For more sophisticated users—for example, the intelligence or defense community—there is a real desire to receive everything contained in all of the messages. Then the users split into those that want data in near real-time, as a streaming data feed into live operational systems, and those that prefer a more on-demand model. Initially, we focused on the near real-time feed of raw AIS messages into command and control and other systems of the common operating picture (COP) type. That initial market is still growing. We delivered about 4 billion AIS messages in this manner in 2012 to large users, mostly defense and intelligence-based.
However, we see a growing need for ship information for various uses — whether it be commercial uses, environmental uses, search & rescue, that sort of thing — where there is not necessarily the need for raw AIS data, which is perceived to be too complex. They don’t necessarily want to set up huge databases and have massive computing infrastructures and database administrators to handle it and are more likely to have a mainstream GIS platform, or even just a Web browser as a means to display data. Our streaming data feed is probably too much, too granular, and too frequent for them. So, we are very quickly moving into the realm of deriving ship and voyage information from the AIS messages data, to make it more meaningful and useful, and then transmitting this information in a more easily ingestible form to a much wider audience. As a result, we have just announced our first generation of geospatial Web services.
S&S: It sounds as though you will be increasingly moving toward providing Web-based services in addition to, or as alternative to, providing the data.
Stickler: Yes, absolutely. There is a lot that we can do with all of the data in the AIS messages to turn it into information that is more targeted at specific use cases and then deliver that as an OGC-compliant Web service that can be very simply and easily integrated. We are trying to lower the barriers of entry for people to start using this information because traditionally, if you did not understand the AIS messages and you did not have the infrastructure in place to ingest them, then it was a very steep learning curve and quite a large investment to get into this market. We are trying to remove those barriers and make global ship information more available to more people more easily.
S&S: Some port authorities use your system for local traffic information?
Stickler: Port authorities already have notice of arrival systems, for example, and they can use their terrestrial AIS to monitor the shipping around their port. But the challenge with that is that the horizon is the limit. They can only see for 50 nautical miles or so away from the port. By adding our data to the existing system, we extend their reach. Now they can start to plan and schedule port arrivals and their port management at a much earlier stage, before the ship gets too close. We’ve seen some very interesting applications, recently, that extend the “notice of arrival” management process.
S&S: What are some other interesting and creative uses of your system?
Stickler: Another interesting application is casualty alerting. AIS data includes such things as the ship’s course over ground, heading, and speed. For a ship, the course over ground and the heading are almost never the same, because ships drift in the currents. But if the course over ground and the heading get wildly different—for example, they are at 90 degrees to each other—that means that the ship is moving sideways. If the speed is very low at that point and the seas are very high, you have a potential problem. If you get a big tanker side on in a huge swell, you could be heading for an environmental disaster. We’ve been working with some authorities to give them an early alert of where a ship may be getting into trouble. So, rather than doing analysis from the past, we are actually looking at predictive or alerting capabilities.
We also have interesting applications in the Arctic. As we all know from the news, Arctic ice is retreating at a rapid rate every year, which is opening up the Arctic to shipping. We have fantastic coverage of the Arctic region from our polar orbiting satellites and have now built up an archive with three years of data, which is invaluable for studying the effects of shipping in this fragile region, and we’ve been using our data to map the way that the shipping is increasing year by year in the Arctic region. The increase in shipping in the Arctic is not just an environmental issue but it also has a major impact on the world’s economy, because using the Northwest Passage, for example, as a route from the Pacific to the Atlantic can save a huge amount of fuel. So, it is a very attractive option.
The issue here though is that there is no infrastructure and the area is very poorly surveyed because there has never been a need to survey it in any great detail. As a result, there are dangers. If a ship gets into trouble up there, it is a long way from any help and the consequences could be severe. This year, for the first time a tanker went across the Russian northern passage during winter. That could have huge implications. We tracked that voyage. We can also count ships going through particular waypoints in the Arctic during the summer, we can do density maps, we can monitor what the passenger ships are doing—the tourist trade is obviously trying to take people to more and more remote new places. That sort of information is very important and interesting to a wide range of authorities.
We’ve also been doing some work with organizations looking at the piracy threat. Pirates tend to take a ship and then use it as the mother ship to launch further attacks. They very rarely, if ever, switch the AIS off because you wouldn’t want to go into a very busy shipping lane—for example, around the Horn of Africa—in a ship with the AIS switched off and being both blind and invisible to the ships around you. So, we can help provide information on these mother ships and their movements.
S&S: What about environmental uses?
Stickler: With our global coverage we can give an unprecedented view of shipping in any area of the world. Remote regions are often those we wish to protect the most. We are able to provide information on when ships are either approaching a marine protected or environmentally sensitive area or can provide statistics on how many go through what types. Often, a marine protected area doesn’t mean that no shipping can go through; it might mean that only ships of a certain type or size are permitted. We can monitor this from space and provide unique insight, as often these marine protected areas are out in the deep oceans. As we have a detailed record of a ship’s voyage there is also a potential to monitor for illegal activity, such as oil bunkering or where they flush their tanks out at sea, making sure that they are carrying out such activities in the right areas, far enough away from the coast line and not close to marine protected areas or environmentally sensitive areas. Certainly we also see a growing interest now in using Satellite AIS to also help with the growing illegal fishing problems.
S&S: What are some examples of how you customize information feeds for customers? Of course, much of this may become moot as you move increasingly toward Web-based services.
Stickler: Some people only want to monitor certain areas of the ocean, some people are only interested in tracking certain ships, and some are interested in a combination of the two. We have various ways to customize what areas of the world or what particular ships they want information on. Obviously, when we get into the Web services world the whole thing becomes a lot more flexible. You can ask for all sorts of combinations of customizations. So, you can say, for example, I want all crude oil tankers that are within 200 miles of the Canadian coast that are doing less than 5 knots and are flagged to the Marshall Islands. You can get very granular with the questions that you ask. Web services is a question and answer, on-demand, model and you only get back the data that you request.
S&S: While it is very dangerous to shut off AIS in high traffic areas, on the high seas people engaged in piracy, illegal fishing, and other criminal activities often do shut off AIS to avoid detection. However, those ships can be spotted using satellite-based radar. Have you done any work to match them up with their AIS tracks?
Stickler: Yes, that’s used a lot. Our data is used extensively in the defense and intelligence arena by fusing it with both radar and optical data. Typically, the AIS is used to effectively clear out the “friendly” ships and only leave behind those ships that aren’t transmitting their AIS. You then immediately know where to focus your resources and your assets. So, you use AIS to eliminate those targets of your radar image—because your radar image will give you a true picture of everything that is on the ocean surface—and you use AIS to eliminate what you know and you are left with the unknowns, which is where you focus your assets.
We are working with the Europeans on a satellite called PAZ. We are going to have one of our AIS receivers on a radar satellite for the first time. Earlier this month, the Canadians announced that they are going ahead with three more satellites in their Radarsat constellation mission and each one of them will also have an AIS receiver on board. That will give users for the first time the ability to capture AIS data at exactly the same time as a radar image. At the moment, the radar image is taken from one satellite, the AIS data comes from another satellite, and they are not coincident in terms of time of acquisition. In this case, we have to do some hindcasting or forecasting and interpolation of positions to best match the AIS signals with the image you see. When these new satellites go up, users will be getting AIS signals at exactly the same time as the radar images are acquired, so that will be quite a significant step forward in that arena.
S&S: Is anyone using your data to study commercial flows?
Stickler: Yes, people do that. Up to this point, we have not really engaged the commercial market directly. We have a partner, VesselTracker, in Germany, that does operate in the commercial market. They are the leading terrestrial AIS provider. They fuse our data with their own data and supply that to ports and other commercial entities, owners, insurers, etc. From a commercial perspective you do get information in the AIS messages about the cargo and destination port and estimated arrival data all of which have a potential commercial ‘value’, but it has to be entered into the system by the people on board the ships, so you have to be a little bit cautious, because what you are getting is not an automatic feed. In shipping around the Horn of Africa, in the piracy regions, there has been quite a lot of use of the phrase “armed guards on board” in the cargo field in the AIS message, just as a warning to all those people in the vicinity, ‘don’t come near me!’
S&S: Have you faced resentment by ship owners, governments, or anybody else about the availability of this data?
Stickler: We operate under a license from the Canadian government, which is a member of the IMO. We simply cannot just supply our data to anyone. The infrastructure and the technology for this system are part of the SOLAS convention, which is all about the safety of life at sea. We may not, under any circumstances, make our data available to anyone if it may end up putting ships and their crews at risk. So, for example, we do not supply data to any environmental groups that may have the dubious tag of being “ecoterrorists.” If this data could be used to endanger lives at sea, then it cannot be released.
S&S: If two ships are very close to each other, side by side, can you still tell which is which?
Stickler: Yes, we can. The density of shipping does have an effect, for sure, but AIS is a very clever, self-regulating system. Ships transmit their data at a certain time and then reserve a time slot for the next one. Within a cell, no two ships transmit at the same time on the same frequency. Those cells are constrained by the horizon—a radio wave can’t be seen over the horizon, so the cell can only be a certain size for that reason—and the cell can get smaller as more ships appear within the vicinity of each other, because otherwise you just don’t have enough time slots for them all to transmit uniquely.
The problem that we have, and anyone looking from space has, is that when you look down you see multiple cells. For us for example, on average we see about 3,000 cells within a satellite footprint. So, in principle, there could be 3,000 ships down there, transmitting at exactly the same time and exactly on the same frequency. We call that “signal collision.” Obviously, the denser an area the more chances of collision you get. Our IP as a company is in the fact that we have written de-collision algorithms in our processing, to de-collide out these signals, so that we can actually get down to picking out each individual ship, regardless of how dense the area is. One of our unique capabilities is that in any one satellite pass, we will get more ships, particularly in very dense areas, than anyone else, because we have this unique de-collision technology.
S&S: What is the importance of open standards in this area?
Stickler: We realized very quickly that we are in the geospatial data business. Interoperability is a key concept for us—we want to get our data out in such a way that people can very easily use it. So, we decided to follow the OGC standards. Our geospatial Web services—both Web mapping services and Web feature services (pictures and data)—are available now, certified to the OGC standards, which means that you can effectively pick up any geospatial client that is OGC-compliant, which is pretty much all of them, and immediately ingest our data. So, whether it is an Esri platform, Google Earth, or any of the free viewers out there, you can open it up and immediately manipulate and view our data. It does not require any coding. That for us was really, really important.
