How radar-assisted situational awareness can reduce risks and improve efficiency in the maritime industry.
Currently, human error accounts for 75–96% of accidents in the maritime industry. With a view to reducing this figure and improving safety, there is an industry-wide push for the automation of many shipping processes. For example, in 2021, the Ever Given blocked the Suez Canal for six days when human error caused the ship to ground. An estimated $9.6 billion in world trade was lost. But would this disastrous event have happened if the Ever Given had navigated autonomously?
Situational awareness – an understanding of where a vessel is and what is around it – is vital for autonomous vessels, which need it to navigate any environment, determine the most efficient route, and avoid harm. While large ships can be detected by traditional marine radars, wherever navigation is autonomous, small objects must also be detected. Small objects commonly found in a marine environment include boats, kayakers, animals, swimmers, and buoys.
At the moment, autonomous and semi-autonomous vessels achieve situational awareness via onboard sensors, such as cameras and lidar. Both of these can provide detailed information about the vessel’s environment, but only within certain operating parameters. Both cameras and lidar have limitations and leave vulnerabilities in terms of situational awareness. Both have limited detection ranges and may struggle with adverse weather conditions, such as fog and rain, and environmental interference, including glare from light reflecting from the water’s surface, or from the sun low in the sky.
Radar has a long history within the maritime industry: X-band (8–12 GHz) radars have been used on boats for decades, to measure bearing and distance from ships to prevent collisions. Marine radars are valuable for their impressive range and ability to perform in very harsh conditions. However, conventional X-band radar lacks the resolution required for use alone in an autonomous vessel, and so developers are looking to W-band (76–77GHz) radar that boasts higher resolution.
Standard marine radars are typically used to detect large objects, such as ships and coastlines. They do, however, have trouble detecting objects smaller than a few metres in size and at close range.
W-band radar is a novel sensing technique that combines the range and reliability of typical X-band marine radars with the accurate data yielded by alternative sensor solutions, such as lidar and camera-based systems. Navtech Radar employs its patented W-band radar technology as part of their market-leading solutions. With wavelengths of a few millimetres, Navtech’s sensors can achieve superb resolution, allowing developers to detect small objects at ranges of up to 500m. Radar returns are simultaneously data-efficient and information-rich: radar can observe multiple objects per transmission and detect their locations and cross-section characteristics. Navtech has been working for the last 20 years to ensure this technology is compact and affordable.
Navtech radars can be operated in a number of different configurations, including a high-precision mode, where objects as small as 4.4cm in size can be detected. Internal trials have validated the detection of kayaks, paddleboarders, and swimmers with a Navtech sensor from onboard a vessel.
Navtech radars provide:
Even where autonomy is minimal, being able to understand the vessel’s local environment is extremely useful, to reduce the risk of collision. Navtech Radar complements typical marine radar with detection capability that finds smaller objects at shorter ranges (less than 500m). It does this while in all weather and lighting conditions – factors that limit the use of both camera and Lidar in marine autonomy.
Navtech ran an internal trial to assess the object detection capabilities of their sensors. The trial was conducted on the River Torridge in Devon, UK. The setup for the vessel used in the trial is shown in Figure 1.
Four objects were chosen to assess detection capabilities. These were a swimmer, a kayaker, a paddleboarder and a buoy – all objects commonly encountered in ports or inland waterways. The ability of the radar to detect a specific object was assessed by looking at its signature on the radar data output. This was then compared to video, captured from an onboard GoPro, and drone footage from above the vessel.
Figures 2 to 5 show the results from this trial. The buoy was detected at a range of up to 300m; the swimmer and kayaker were detected by the radar at a range of up to 250m.
Navtech ran an internal trial to assess the object detection capabilities of their sensors. The trial was conducted on the River Torridge in Devon, UK. The setup for the vessel used in the trial is shown in Figure 1 (left).
Four objects were chosen to assess detection capabilities. These were a swimmer, a kayaker, a paddleboarder and a buoy – all objects commonly encountered in ports or inland waterways. The ability of the radar to detect a specific object was assessed by looking at its signature on the radar data output. This was then compared to video, captured from an onboard GoPro, and drone footage from above the vessel.
Figures 2 to 5 show the results from this trial. The buoy was detected at a range of up to 300m; the swimmer and kayaker were detected by the radar at a range of up to 250m.
Figure 2: radar data overlaid onto an image at the same instance, taken from a drone directly above the vessel. This clearly demonstrates detection of a swimmer (right) and kayaker (left)
Figure 3: radar data overlaid onto an image at the same instance, taken from a drone directly above the vessel. Detection of a paddleboarder (left) and kayaker (right) is clearly shown
Radar data (top) showing detection of kayaker (circled in red) from a range of around 250m from the vessel. Image (bottom) taken at the same instance from an onboard GoPro.
Radar data (top) shows detection of a buoy (circled in red) around 50m from the vessel. The image (bottom) was taken at the same instance from an onboard GoPro.
The trial results clearly show that a Navtech radar sensor can be used effectively for small object detection. A sensor with such a range, and ability to detect frequently-encountered small objects such as swimmers, can be used to enable situational awareness in a range of autonomous marine projects.