GPS World

Tag: self-driving

  • Cleared for the dirt: How robotic rovers are revolutionizing military runway assessment

    Cleared for the dirt: How robotic rovers are revolutionizing military runway assessment

    Tactical air-lifters such as the Airbus A400M, Lockheed C-130 and Boeing C-17 require precise runway roughness assessments to operate safely on unpaved surfaces. An autonomous rover system developed at the Royal Military Academy of Belgium uses RTK/PPK GNSS positioning and sensor fusion to deliver centimeter-level height measurements, drastically reducing survey time. The system provides a practical solution for rapid runway certification across military operations and humanitarian response missions.

    Unpaved runway assessment

    The Airbus A400M Atlas, the Lockheed C-130 Hercules and the Boeing C-17 Globemaster III routinely operate from unpaved runways in harsh environments far from established infrastructure. Before these aircraft can safely land, flight crews require accurate runway roughness data to assess whether the surface meets operational limits. This assessment relies on precise, quantitative measurements of the runway’s surface characteristics — a task that traditionally requires specialized survey teams and hours of manual work with GNSS equipment, resources that are often unavailable in high-tempo tactical or emergency response scenarios.

    The challenge is particularly acute because different aircraft have specific roughness tolerances. The A400M uses an equivalent bump height (EBH) methodology, while Boeing employs its Boeing Bump Criteria. The EBH requires vertical measurement precision of ±1 cm over wavelengths ranging from 5 to 100 meters. Meeting these stringent requirements with rapid, field-deployable methods has remained an operational gap — until now.

    At the Royal Military Academy (RMA) of Belgium, we developed a novel solution to this critical challenge. Our system features a rugged, autonomous unmanned ground vehicle that can rapidly perform a centimeter-accurate runway assessment with minimal user intervention. It represents a fusion of robotics, geodesy, and advanced GNSS techniques, designed specifically for ease of use by military teams in the field. The system is called Belgian Navigational Surface Inspector (BENSI).

    FIGURE 1 shows the BENSI system during a mission at a tactical landing zone with the A400Min the background. FIGURE 2 shows the BENSI system being configured by the operator during a landing preparation.

    Figure 1 The autonomous UGV (BENSI) during a mission at a tactical landing zone with the A400M Atlas in the background.
    Figure 1 The autonomous UGV (BENSI) during a mission at a tactical landing zone with the A400M Atlas in the background.
    Figure 2 The BENSI system being configured by the operator during the beach landing preparation at Rømø, Denmark.
    Figure 2 The BENSI system being configured by the operator
    during the beach landing preparation at Rømø, Denmark.

    This article details the system’s architecture, the integration of multiple technologies that enable the stringent precision required achieved by GNSS and sensor fusion, self-driving capabilities and its successful deployment in demanding field tests. We present a military graded solution for ensuring tactical airlift safety, enabled by modern, accessible GNSS technology and robotics.

    Quantifying runway roughness

    Deployable Air Traffic Management (DATM) and Pathfinders are responsible for ensuring the safety of aircraft operations on unpaved runways. They are tasked with assessing the quality of the runway and the Runway Safety Area (RSA) to ensure that the aircraft can land safely. The pilots analyze their assessment and take the final decision to land.

    FIGURE 3 is an example of a landing zone having an unpaved runway that needs to be evaluated for landing. FIGURE 4 overviews the landing zone by mapping and indicating features of the runway that need to be considered by the pilots. An important aspect of the DATM’s assessment is the runway’s roughness, which is quantified by the EBH.

    Figure 3 An example of a tactical landing zone.
    Figure 3 An example of a tactical landing zone.

    For modern military transport aircraft operations, runway roughness assessment is a critical safety parameter. Both major manufacturers — Airbus with its EBH methodology and Boeing with its Boeing Bump Criteria — have developed sophisticated approaches to characterize runway longitudinal roughness profiles. These methods analyze height variations over wavelengths ranging from 5 to 100 meters, requiring vertical measurement precision of ±1 cm. This rigorous assessment is essential to reduce aircraft structural fatigue, minimize maintenance costs, prevent exceedance of design limit loads, and ultimately ensure safe operations. For the A400M specifically, Airbus requires EBH characterization to determine operational limitations of the aircraft’s maximum payload.

    Figure 4 A typical mapping of a landing zone showing a condensed overview of DATM’s assessment.
    Figure 4 A typical mapping of a landing zone showing a
    condensed overview of DATM’s assessment.

    Traditionally, achieving this precision would involve a painstaking survey conducted by specialists using a GNSS survey system mounted on a trolley requiring human guidance along the measurement tracks totaling more than 3 km of length. For military units like the DATM and Pathfinder teams, who often are the first on the ground, this is impractical. They need a system that is rapid, reliable, simple to operate without a surveying background, and robust enough for field conditions.

    A GNSS-Centric design

    Our solution is a two-part system designed for rapid deployment: a portable GNSS base station and autonomous rover. FIGURE 5 shows a schematic overview of the system architecture.

    Figure 5 A schematic overview of the system architecture, showing the data (NMEA) and correction (RTCM) flow between the base station, rover and operator.
    Figure 5 A schematic overview of the system architecture, showing the data (NMEA) and correction (RTCM) flow between the base station, rover and operator.

    The base station: The system’s anchor

    Housed in a compact, portable case, weighing just 2 kg including tripod and radios (as seen in FIGURE 2), it serves as the operational hub. Once set up on its lightweight tripod, it performs an automatic survey to establish its precise coordinates. Its primary role for positioning is to generate and transmit Radio Technical Commission for Maritime Services (RTCM) 3.x correction data to the rover via a robust long-range radio link (operating in the868/900MHz bands).

    Beyond its GNSS duties, the base station acts as a self-contained command center. It hosts a Wi-Fi hotspot and a web server, allowing the operator to connect with any standard tablet, smartphone or laptop. This web interface is used for mission planning, command and control of the rover, and real-time monitoring of survey progress. At the end of the mission, the operator can download the EBH data and additional quality metrics of the runway for analysis such as a summary report of the complete measurement, a gradient analysis, and a runway map highlighting zones with bumps or troughs exceeding the specified criteria.

    An autonomous, all-terrain surveyor

    The UGV is a lightweight but rugged platform chosen for its durability and open-source software architecture, which allows for deep integration of our custom navigation and control algorithms. The rover has been designed to be able to traverse rough terrain and survive in harsh weather conditions. The UGV consists of two parts, the chassis (11 kg) and the processing payload(8 kg). The heart of the rover is the processing payload, which contains a sophisticated sensor suite designed for high-precision localization and navigation.

    ■ Primary GNSS receiver. A high-grade, multi-constellation Septentrio receiver with a Calian/Tallysman GNSS antenna provides the main source of positioning information.

    ■ GNSS heading. A second Calian/Tallysman GNSS antenna, set up in a moving-base configuration, provides degree-accurate true heading, which is critical for maintaining precise track-following.

    ■ Inertial measurement unit (IMU). An industrial-grade Xsens IMU provides high-frequency data on the rover’s orientation and acceleration, bridging any brief GNSS outages, providing the sensor fusion algorithm with high-rate data, and helping to smooth the final trajectory.

    ■  Radio communication. The radio modules provide robust long-range communication with the base station operating in the 868/900MHz bands.

    ■ Wheel odometry. Encoders on the rover’s wheels provide continuous velocity information, acting as a crucial input for the sensor fusion algorithm. All sensor data is fed into an onboard mini-PC running the Robot Operating System, a flexible framework for developing robotic applications.

    Path to precision

    Achieving centimeter-level accuracy on a moving platform in challenging environments requires more than just a good GNSS receiver. Our approach is built on a robust foundation of sensor fusion and a dual processing strategy using real-time kinematic and post-processing kinematic (RTK/PPK). An extended Kalman filter (EKF) is at the core of the rover’s navigation software. The EKF continuously fuses data from the GNSS receivers, IMU and wheel encoders to produce a single, high-integrity “pose” (position and orientation) estimate.

    For runway surveying, we employ two modes of GNSS processing:

    RTK. During the mission, the rover uses the RTCM corrections from the base station to compute a centimeter-accurate position in real-time. This is used for autonomous navigation, allowing the rover to follow its generated mission plan configured by the operator with high precision.

    PPK. While RTK provides excellent real-time results, the most demanding applications benefit from post-processing. Both the rover and the base station log all raw GNSS observables during the mission. After the survey is complete, these raw data files are processed together which allows for more rigorous quality control and can often resolve ambiguities or fix cycle slips that were not solvable in real-time, providing the definitive, highest accuracy trajectory for the EBH analysis.

    A final crucial step is extracting the height profile for each EBH track and subsequently transforming and reformatting this data for Airbus’ AssurTool. The step also is automated and carried out by the software. It takes care of the following:

    ■ The conversion of the geodetic coordinates (latitude, longitude, and height above the World Geodetic System 1984 [WGS84] ellipsoid) to Universal Transverse Mercator plane coordinates and orthometric heights (heights relative to a geoid).

    ■ The extraction of the height profile of each EBH track.

    ■ Quality control of the precision of the height profile flags tracks that do not meet the required accuracy or show inconsistencies.

    ■ The transformation and reformatting of this data for Airbus’ AssurTool.

    Self-driving capabilities

    The rover uses a navigation framework with a custom planner for generating smooth, curved paths that match the rover’s turning capabilities and steers the rover using a controller based on the Regulated Pure Pursuit tracking algorithm. A specialized lane-generation algorithm creates optimal survey patterns from runway corner points, with behavior-tree recovery strategies for robust operation.

    FIGURE 6 shows a typical EBH survey pattern generated from the mission plan and executed by the rover and a depiction of how the rover plans the smooth curved path between the lanes.

    Figure 6 Features of the navigation framework used for planning the EBH tracks. (a) A typical EBH survey pattern generated from the mission plan and executed by the rover. (b) A depiction of how the rover plans the smooth curved path between the lanes.
    Figure 6 Features of the navigation framework used for planning the EBH tracks. (a) A typical EBH survey pattern generated from the mission plan and executed by the rover. (b) A depiction of how the rover plans the smooth curved path between the lanes.

    A streamlined workflow

    The system was designed from the ground up to be operated by non-surveyors. A typical mission workflow is as follows:

    Setup. The operator places the base station on a tripod near the runway and unfolds the rover. The entire hardware setup takes less than 10 minutes.

    Mission planning. Using a ruggedized tablet (or any other device with a web browser), the operator connects to the base station’s WiFi and opens the web interface. They define the runway by entering the coordinates of the runway’s corners. The software automatically calculates the EBH lines based on the required spacing. FIGURE 7a shows the user interface displayed on a tablet, showing the EBH mission configuration page.

    Figure 7a The user interface of the web application.
    Figure 7a The user interface displayed on a tablet, showing the EBH mission configuration.

    Execution. The operator initiates the mission, and the UGV autonomously navigates to the start of the first line and begins the survey. The operator can monitor and control the rover’s progress, position, and GNSS quality status in real-time on the web interface. FIGURE 7b shows the user interface displayed on a tablet, showing the rover control, the real-time status of the UGV and the measurements.

    Figure 7b The tablet showing the rover control and the real-time status of the UGV and the EBH results.
    Figure 7b The tablet showing the rover control and the real-time status of the UGV and the EBH results.

    Data retrieval. Upon completion, the rover returns to the base station. The system automatically processes the data, producing downloadable files formatted for direct import into Airbus’ AssurTool and additional useful quality metrics for the operator. These consist of a summary report of the complete measurement, a gradient analysis, and a runway map highlighting zones with bumps or troughs exceeding the specified criteria.

    Analyzing the data

    Once the rover completes its survey and returns to the base station, the system automatically initiates post-processing of the collected data. This critical step validates the quality of every measurement and generates operator-ready outputs for both Airbus’ AssurTool and field assessment.

    The post-processing pipeline applies rigorous quality criteria to each survey line. Lines failing these criteria are automatically flagged with detailed diagnostics explaining the cause.

    For operational decision-making, the system generates a comprehensive visualization report. The operators receive planimetric maps showing the height profile plots and a detailed gradient analysis identifying critical slope transitions. A key capability is the generation of a 3D interpolated height map of the entire runway surface. This color-coded surface map provides an intuitive view of the runway’s topography, clearly highlighting zones with excessive bumps, depressions, or gradient anomalies that facilitates the assessment of the runway.

    These analysis reports are accessible through the web interface for immediate download to the operator’s tablet. FIGURES 8 shows examples of the visualization report.

    Figure 8a 2D height and gradient contour maps of two surfaces generated by the BENSI system. (a) A height contour map of two landing zone (LZ) surfaces automatically generated by the BENSI system.
    Figure 8a 2D height and gradient contour maps of two surfaces generated by the BENSI system. (a) A height contour map of two landing zone (LZ) surfaces automatically generated by the BENSI system.
    Figure 8b A gradient contour map of two LZ surfaces automatically generated by the BENSI system.
    Figure 8b A gradient contour map of two LZ surfaces automatically generated by the BENSI system.

    Proven performance

    The UGV system is a mature prototype that has been validated in numerous international military exercises. It has successfully surveyed tactical landing zones in varied environments, from the desert strips of Yuma, Arizona, and 29 Palms, California, to the sandy shores of Denmark and fields in France, Portugal and Italy. In all tests, the system has consistently delivered the sub-centimeter height precision required for A400M EBH certification.

    2025 Rømø Head-to-Head Trial. During beach-landing preparations in August 2025, our autonomous rover and a manual system (human-guided trolley) using a professional GNSS survey system ran side-by-side on a 1 000m landing zone on the Rømø beach in Denmark. The BENSI solution matched the manual survey system height profile with a standard deviation of 8mm and demonstrated significantly better lane-tracking consistency (mean deviation: 8,5 cm vs 16 cm and deviation error: 3 cm vs 9 cm). FIGURE 9 shows the height-error distribution between the BENSI system and the manual survey system at Rømø, Denmark.

    Figure 9 Height-error distribution between the BENSI system and the manual survey system at Rømø, Denmark.
    Figure 9 Height-error distribution between the BENSI system and the manual survey system at Rømø, Denmark.

    Rapid humanitarian response

    While BENSI was conceived for tactical airlift operations, its capabilities extend naturally to humanitarian assistance and disaster-relief missions. Belgium’s civil rapid-response unit Belgian First Aid & Support Team (B-FAST) routinely deploys doctors, paramedics, firefighters, and other professionals worldwide following earthquakes, floods, or epidemics. Leveraging the A400M’s ability to land on short, unpaved strips away from congested or contested airfields drastically cuts transit times — but only if the runway’s condition can be certified quickly.

    The BENSI systems enables a DATM team to quickly relay an EBH report and awareness map of the immediate area to the inbound aircrew. This rapid assessment unlocks critical early access for life-saving medical supplies and personnel when every hour counts.

    Conclusion and the Road Ahead

    The fusion of autonomous robotics and high-precision GNSS offers a powerful solution to the critical challenge of certifying unpaved runways. Our system saves valuable time, reduces the burden on specialized personnel, and provides objective, high-quality data that directly enhances the safety of tactical airlift operations.

    Development is ongoing. Our current efforts focus on several key areas:

    ■  Improving navigation in degraded environments. We are exploring tighter coupling between the GNSS and IMU to provide more robust navigation through areas of poor satellite visibility.

    ■ RSA assessment. We are experimenting with integrating a lidar sensor to generate a 3D point cloud of the runway and its surroundings. This will automate obstacle detection and the assessment of the RSA, though we are carefully working to mitigate potential electromagnetic interference from the lidar that can interfere with GNSS reception.

    ■ Handheld corner point device. To further improve absolute accuracy, we are developing a small, handheld device that uses RTK corrections from the base station, allowing operators to mark the runway corners with centimeter-level precision.

    This project demonstrates a clear application of GNSS technology in a demanding military aviation context, with broader implications for any field requiring rapid and precise surface profiling, from civil engineering to disaster response.

    Development Team

    ■ Pieterjan De Meulemeester ([email protected]) is a Ph.D. research engineer at the RMA of Belgium.

    ■ Alain Muls ([email protected]) is professor at the RMA of Belgium. He teaches the courses Military Satellite Based Positioning andMilitary Geodesy.

    ■ Jarno Van Audenhoven ([email protected]) is a Robotics Development and Research Engineer at the RMA of Belgium.

    ■ Pascal De Kimpe is a technician at the RMA of Belgium.

    ■ The BENSI system was developed by the R&D team at the RMA of Belgium in collaboration with Belgian Defense. The system has been successfully field-tested during international military exercises and is being evaluated for operational deployment.

    All photos courtesy of BENSI Development Team of the Royal Military Academy of Belgium

  • Virtual vineyards created for self-driving tractors

    Virtual vineyards created for self-driving tractors

    While grapes are being harvested throughout Italy, the Politecnico di Milano is looking to the future of viticulture with an innovative approach that combines mechanics, IT and digital simulation.

    A team of researchers from the Departments of Mechanical Engineering and Electronics, Information and Bioengineering at the Politecnico di Milano has developed a system to test and optimize self-driving strategies for agricultural tractors in a virtual environment.

    The study, published in AgriEngineering (“Scenario Generation and Autonomous Control for High-Precision Vineyard Operations}, presents a complete methodology for creating realistic vineyard scenarios and evaluating control algorithms for autonomous driving. The goal is not simply to reduce the human presence, but to provide a high-fidelity digital environment in which to develop, verify and safely improve agricultural automation solutions based on sensors and predictive algorithms.

    The research has made it possible to create a digital twin of the vineyard, capable of reproducing slopes, soil irregularities and row layout. Tractors equipped with low-cost GNSS and inertial measurement systems (IMS) sensors and guided by advanced algorithms have been tested in this virtual environment, vehicles capable of moving autonomously between rows and of performing off-field turning manoeuvres with the utmost precision.

    The study explored new methodologies to simulate and independently control vineyard operations. (Credit: Politecnico di Milano, CC BY-SA).
    The study explored new methodologies to simulate and independently control vineyard operations. (Credit: Politecnico di Milano, CC BY-SA).

    “Our approach combines terrain modeling, advanced control and realistic sensors in a single simulation environment. This speeds up research and reduces the risks and costs of real field tests,” said Federico Cheli, professor at the Politecnico di Milano, Department of Mechanical Engineering, and project coordinator.

    According to the researchers, the use of realistic simulations not only reduces the risks and costs of field tests, but can also become a useful tool for operator training. It can accelerate the adoption of new agricultural technologies.

    The project stems from the partnership between researchers at the Politecnico di Milano and the company Soluzioni Ingegneria s.r.l. that develops software for dynamic vehicle simulation. It is part of a broader context of cooperation with industrial companies engaged in research on automation and sustainability in agriculture.

    Ruiz Mayo, C.; Cheli, F.; Arrigoni, S.; Paparazzo, F.; Mentasti, S.; Pezzola, M.E. Scenario Generation and Autonomous Control for High-Precision Vineyard Operations. AgriEngineering 2025, 7(2), 46. https://doi.org/10.3390/agriengineering7020046

  • SBG Systems drives GNSS+inertial in Paris

    SBG Systems drives GNSS+inertial in Paris

    Photo: SBG SystemsAutonomous vehicles require lane-level accuracy at all times and in all conditions. However, under many conditions, such as in urban canyons and tunnels, they may lose line-of-sight to enough GNSS satellites to achieve accurate and robust positioning or may have no signal at all. In these situations, they need data from other sensors, including an odometer and an inertial measurement unit (IMU). Creating reliable and safe autonomous navigation requires fusing GNSS and inertial technology in a multi-layered system.

    SBG Systems and its partners LeoDrive.ai and Intempora, have been doing this to develop solutions for autonomous vehicles. SBG’s technology enables multi-sensor integration while addressing such autonomous navigation challenges as time synchronization, integrity, precise positioning and high-definition mapping.

    “To ensure performance and build trust, we assemble our own IMUs from carefully selected industrial-grade parts, then we calibrate all our products individually,” said Laurent Le Thuant, business manager for SBG, in a recent webinar.

    For safe operation, Le Thuant explained, the vehicle’s true positional error (PE) must be smaller than its protection level (PL), which in turn must be smaller than its alert limit (AL): PE < PL < AL. Otherwise, the solution is declared unavailable or reports misleading information.

    In automotive tests conducted in a business district near Paris, an SBG vehicle was equipped with both a GNSS-only, automotive-grade multiband RTK receiver equipped with a PL determination algorithm and an RTK GNSS receiver tightly-coupled with an IMU and an odometry input. A comparison showed that the former was not suited for self-driving, while the latter significantly improved the solution availability, accuracy and protection levels.

    For self-driving in the most severe conditions, even this solution requires integration of supplementary sensors, such as cameras, lidars and radars for precise localization.

  • Domino’s delivers with Nuro and GNSS

    Domino’s delivers with Nuro and GNSS

    Photo: Domino's
    Photo: Domino’s

    In April, the pizza company Domino’s and self-driving delivery company Nuro launched autonomous pizza delivery in Houston, Texas. Select customers who place a prepaid online order on certain days and times from Domino’s in Woodland Heights can choose to have their pizza delivered by Nuro’s R2 autonomous, occupantless on-road delivery vehicle.

    Customers selected for the service receive text alerts, which update them on R2’s location and provide them with a unique PIN to retrieve their order. Once R2 arrives, customers are prompted to enter their PIN on a touchscreen, opening its doors.

    In February 2020, Nuro became the first autonomous vehicle developer to be given exemptions by the U.S. National Highway Traffic Safety Administration for testing on public roads without the need to have controls for human operators. Unlike many other autonomous vehicle companies, Nuro engineered its self-driving road vehicles to transport goods instead of people.

    There’s no set timetable for how quickly Domino’s and Nuro will evaluate their testing or expand the service.

    Nuro is also carrying out trials and pilot deliveries with several other companies, including restaurant chain Chipotle, Kroger grocery stores, CVS pharmacies, Walmart and FedEx.

  • The trolley problem: What would a self-driving car do?

    The trolley problem: What would a self-driving car do?

    Image: metamorworks/iStock/Getty Images Plus/Getty Images
    Image: metamorworks/iStock/Getty Images Plus/Getty Images

    Years ago, a trucker driving down the western slope of the Rocky Mountains lost his brakes. As his truck accelerated, he hoped to make it to the next runaway truck ramp before losing control. However, when he reached it, he saw a car parked at its base with a group of teenagers drinking beers. In a split-second decision, he veered to the left instead and went off the cliff. In the coming years, faced with the same moral dilemma, what would a self-driving truck do?

    Matteo Luccio
    Matteo Luccio

    Many similar scenarios have been discussed in the technical literature on self-driving vehicles. Most of them are variations on the “trolley problem” presented to generations of college philosophy students since it was first formulated by philosopher Philippa Foot in 1967 and adapted by Judith Jarvis Thomson in 1985. In the trolley problem, a person can choose to divert a trolley from the main track, saving five people who are working on it but killing a person on the other track who otherwise would not have been involved.

    When faced with an inevitable crash, should a self-driving car slam into a wall to save the lives of three children crossing the street or, in effect, target them to save its two occupants? Most people, when polled, choose the former. When shopping for a new car, however, those same people are more likely to buy one that will make their own safety its highest priority.

    Human drivers react to emergencies instinctively — motivated by neither forethought nor malice — and in real time. By contrast, the choices made by autonomous vehicles are predetermined by programmers; their control systems can potentially estimate the outcome of various options within milliseconds and take actions that factor in an extensive body of research, debate and legislation. Therefore, our judgment is harsh if those vehicles make what we deem to be the “wrong” choice.

    However, there is no universal agreement as to what constitutes the “right” choice, other than the fact that people generally prefer self-driving cars to minimize the number of lost lives and to privilege people over animals and younger people over older ones. General principles such as “to minimize harm” are of little help in complex and dynamic real-life situations.

    Self-driving cars, in addition to their many other benefits, will dramatically reduce traffic accidents and fatalities, because they will never be distracted, drowsy, drunk or drugged. Yet accidents will still happen, and their outcomes will be largely determined far in advance.

    The mass introduction of self-driving cars onto public roads will require overcoming technical, legal and ethical challenges. As a society, we will have to agree on a uniform set of ethical codes that will guide these vehicles’ decision-making processes in emergencies. This will force us to explicitly quantify the value of human life and property, and encode it in software. These are hard and uncomfortable choices.

    Autonomous systems, fusing data from multiple sensors, will guide these vehicles. It is up to us to decide whom they will target and whom they will spare.

    Matteo Luccio | Editor-in-Chief
    [email protected]

  • Unmanned and AI: Indy Challenge takes autonomous to big track

    Unmanned and AI: Indy Challenge takes autonomous to big track

    When I saw that there was a plan for a whole bunch of unmanned, semi-autonomous racecars to compete at the Indianapolis Motor Speedway (Indy, or IMS) racetrack, I initially thought we might be headed to one significant mess of broken-up machines and potentially a lot of damage. I tracked the various announcements of the competition as things progressed, especially when a prize of $1 million dollars was put up by the Lilly Endowment in Indianapolis, and the majority of the field appeared to be potentially staffed by undergrad university teams.

    Photo: Indy Autonomous Challenge
    Photo: Indy Autonomous Challenge

    However, this isn’t the first time we’ve had unmanned, autonomous road vehicles in competition — we’ve seen highly instrumented SUVs in desert settings in Nevada and California, initially with pretty poor results, which began to improve significantly for the second time round, then vehicles in some simulated street settings with some mixed and also some pretty good results.

    So, as the competition date grew closer for the Indy Autonomous Challenge (IAC), the number of published progress reports began to increase, and we began to better understand how the initial 40 teams might take on this seemingly impossible task — how on Earth will they replicate a regular Indy (also a class of racecar) race? Surely many unmanned racecars on the same track at the same time doing more than 150 mph would be catastrophic!

    When you take a look, however, at the advances we’ve seen, which have enabled unmanned cars, trucks, taxis and such – surely this tech could stretch to meet these major objectives? But Dallara AV-21 Indy Light racecars avoiding hurtling walls passing by, cornering, getting in and out of the pits, coping with vehicles behind, ahead and overtaking — even a superior-equipped unmanned racecar at >150 mph — well that’s something we would really need to see.

    Then you have to take a look at the outfits involved, providing support to the IAC teams – companies including Cisco, and motor sport units such as ADLINK, Ansys, Aptiv, Bridgestone, Luminar, Microsoft and Valvoline and the non-profit Energy Systems Network. The University teams from around the world themselves appeared to also have significant heritage and skill-levels.

    As the 40 University teams started the long trek to get over the hurdles that this challenge presented, members from 21 of those institutions were actually able to make it to Indy, grouped into nine “national” teams. By October 23 the nine teams, with only one car each, were ready to test their autonomous vehicles on the actual track.

    Clemson University established the baseline Dallara AV-21 vehicle and technology to be used by each team for the race, with sensors monitoring chassis motion, suspension, tires and powertrain. Each team would install its own guidance and avoidance system, with each vehicle equipped with six cameras, four lidars, RTK GNSS, associated radios and bags of computing running each team’s customized control system software. The object being for cars to exit pit-lane, accelerate, brake, establish an optimum line for each corner and flat, avoid obstacles, evaluate the track conditions and establish tolerable limits.

    The teams were required to complete several stages of selection, from submission of initial proposals through demonstration of existing vehicle automation capability, simulated race performance, qualification testing at the Indy track — all leading to an anticipated head-to head race against the other qualifiers.

    Then 20 days of planned testing stretched to 50, and three months of preparation passed with students working intensely throughout, curing the glitches, experimenting with how to increase lap speed, and pushing the limits while still keeping the cars intact.

    Energy Systems Network managed the rules of the final competition in a way that reflected Indy qualification days prior the main race — they judged that the technology was not yet at a stage where multiple cars on the track at the same time would have been such a good idea. So, each car was to individually run a number of practice/qualification laps and the quickest car would be the winner.

    During the first stage of live competition, cars were required to exit the pits and run a warmup lap, followed by two laps that were timed and a slow-down lap that required navigating around inflatable barriers on the front-stretch, and then return successfully back around the track into their pit-stop locations. There were several spins in the corners and several crashes, but the four surviving cars/teams were able to optimistically post speeds of more than 130 mph.

    The winning Technical University of Munich team. (Photo: Indy Autonomous Challenge)
    Photo: Indy Autonomous Challenge

    The final phase involved the four teams taking their cars around a number of warm-up/practice laps, followed by four timed laps. Only the car from Germany’s Technical University of Munich was able to complete all laps with an average speed of ~136 mph, so that team ultimately won the $1 million prize. Even so, all teams were able to successfully mature their systems’ performance through the many months leading up to the IAC and their progress through the various qualification stages. Even the other three final qualifiers had much to celebrate as a result of the competition.

    The sponsors supporting the various teams as they progressed through the Challenge may have spent more than $120 million, so that high-pressure development work will be invested back into many vehicle automation opportunities. After all, that was the main objective for the whole undertaking. We should hopefully begin to see safer, more capable self-driving vehicles emerge in the months to come as the technology is applied to more production vehicle automation.

    Tony Murfin
    GNSS Aerospace

  • Tesla announces 1 billion driverless miles

    “As of today [Nov. 28] Tesla owners have driven 1 billion miles with Autopilot engaged,” the company announced via tweet.

    The Autopilot feature became available in 2015 and now comes  on all new Tesla models with a $5,000 activation fee at the time of purchase or $7,000 if selected later.

    The company is training its “neural networks” to improve its self-driving system.

    Photo: Tesla
    Photo: Tesla

    Tesla’s global fleet totals more than half a million vehicles, and recently marked a 20-billion mile step of total electric miles driven, the company said.

    The Autopilot system can also function in the background of the vehicle, without being activated and with no input on control. Thus it gathers data from many more billions of “drivered” miles about its environment and potential Autopilot behavior.

    The company previously mentioned the 1 billion-mile autonomous mark as the minimum it would need to move Autosteer from beta to a regular feature.

    Updates to Autopilot are planned for 2019, including new hardware that will aid in the rollout of the company’s Full Self-Driving system, possibly by the end of that year.

  • DOT ignoring GPS vulnerabilities — again

    DOT ignoring GPS vulnerabilities — again

    The U.S. Department of Transportation’s most recent document preparing for the future of self-driving cars almost entirely ignores positioning, navigation and timing (PNT) needs, according to the Resilient Navigation and Timing (RNT) Foundation. And when it does address GPS, it gets things wrong. A Dec. 3 deadline looms for interested parties to file their comments with DOT.

    In comments submitted to the department’s docket for “Preparing for the Future of Transportation: Automated Vehicles 3.0,” the Foundation — of which I am president — observes that the document does not address GPS service denial at all. While GPS spoofing is mentioned once, the two activities cited as addressing the problem are not PNT-related efforts.

    The comment period is open until December 3. Interested parties can make their own comments and read those already submitted at the website for Docket DOT-OST-2018-0149.

    The cited comment from the RNT Foundation states that, while most self-driving cars are being designed to navigate without external inputs, GPS/GNSS will still be required to initialize location information for vehicle cold startups. Also, most vehicles will reference GPS/GNSS when communicating their positions to other vehicles and traffic control systems.

    Much of the benefit of automated vehicles will come from their participation in Intelligent Transportations Systems. This means wireless networks. The RNT Foundation also urges the department to consider these networks’ critical dependence on GPS timing synchronization in their plans going forward.

    (Image: Pavel Vinnik/Shutterstock.com)
    (Image: Pavel Vinnik/Shutterstock.com)

    The Secretary of Transportation has had a mandate to provide a backup capability for GPS since 2004 that has not been acted upon. The RNT Foundation comments observe that doing so could greatly mitigate all of the concerns mentioned.

    Dana Goward is president of the Resilient Navigation and Timing Foundation, based in Washington D.C.