Technologies, Tools and Platforms Track

In contrast to traditional industrial and consumer robots, field robotics systems must operate in outdoor, unstructured and dynamic environments, such as construction worksites, mines, and farm fields, as well as busy urban centers. Field robotics systems are typically larger and more ruggedized than systems designed for indoor use, and must be able to withstand challenging working conditions. To do so, field systems must incorporate and utilize motion control and actuation technologies that are up to the task.

In this panel session, the unique motion control requirements for field systems will be reviewed, along with the characteristics and capabilities of the latest motion control products and technologies for field robotics systems. Case studies and product examples will be used to highlight salient points. Topics include:
• Field Robotics Motion Control Requirements
• Ruggedized Motion Control Products and Technologies
• Robotic Motion Control Trends for Field Systems
• Applications and Vertical Markets

Battery decisions for mobile robotics are critical for achieving power autonomy, but they can be complex:
• How much capacity is needed?
• Where can I put the Battery?
• What is the best charge method?

In this session, attendees will learn the answers these questions (and more). In addition, detailed information for designers who need reliable portable power for mobile robotics will be provided. Additional topics covered during the session include system considerations such as non-motive loads, battery placement issues, and how to design for re-generative braking charging. Also discussed will be wired and wireless charging pros and cons, plus the trade-offs between short, quick charging and once-per-shift charging methodologies.

For decades, medium and large size off-road equipment engineering and manufacturing companies, and more recently venture investors, have applied large amounts of resources to develop functional off-road highly automated and autonomous machines in field like agriculture, construction, mining, turf. Yet there are few examples of large-scale commercial success of autonomous vehicles (mining being a notable exception).
Product liability is a key obstacle to this commercial success, due mostly to the absence of a local operator, who traditionally held much of the responsibility. Formalizing liability exposure among the various actors (equipment owner, manufacturer, distribution, insurance, etc.) in the absence of a local operator can be managed through the necessary level of safety. Also, straight adoption of on-road approaches may drive unneeded complexity and cost, as on-road methods and standards are geared towards a fundamentally different risk profile. In addition, off-road solutions on lower sales volumes become impractical. But volume issues can be partially offset by leveraging automotive hardware and software solutions that are carefully tailored.

This presentation examines several dimensions of attaining product safety goals in a way to manage liability and reduce business risk. These dimensions include the tailoring of functional safety standards to objectively characterized risks present in off-road applications, integrating vehicle security as a necessary element of product safety, adoption of Safety Of The Intended Function (SOTIF) standards, and other elements of a verifiable and documented safety case.

While there are numerous opportunities for task automation in unstructured environments, physically traversing these settings can be challenging. Sometimes locomotion difficulties can be solved simply, but in other cases, challenges may require an entire design process dedicated to solving the problem of locomotion.

Questions to ask during the design process include: What is the terrain or surface to be traversed? Are there hazards such as water, mud, or even electromagnetic or radioactive fields? How many degrees of freedom are necessary for navigation and task completion? What load must the locomotion system carry? What level of positional accuracy is needed? With these considerations, and many more in mind, it is important to understand various locomotion methods used for automating tasks in unstructured environments as opposed to simply sticking with a potentially limiting method because it is the known and comfortable option.

Over the years, Southwest Research Institute (SwRI) has utilized numerous locomotion methods in a variety of unstructured environments. In this presentation, SwRI’s Branson Brockschmidt will present case studies for several means of locomotion to describe what the design process is like and how a given method can be decided upon. The case studies presented include omni-directional wheels to drive on aircraft depot tarmacs, treads for navigating uneven metal floors in a nuclear waste storage tank, an eight-armed inspection system to climb power line insulators, a tracked gantry for traversing ship ballast tanks, a foam vacuum roller to climb a concrete dam, and drones for remote inspection.

Field robots are being pushed into unexplored territories with tough challenges in unstructured surroundings such as open field, mines, construction sites, and underwater. Truly mastering these critical applications requires additional considerations for difficult conditions that robots must survive. This presentation will answer three critical questions to actuate your robots in severe environments.
• What are the special considerations for robots operating in harsh and variable environments?
• How can my robot survive in their harsh and variable conditions?
• How do I design a robust system that can survive extreme environments?

If you are a field robotics design engineer taking on the challenge of unexplored territories, or an operational decision-maker looking to gain a better understanding of critical robotic components, this presentation is for you.

Using simulation to develop field robotics systems can feel like a boondoggle, especially when you build in sim does not accurately translate to the operational version. It is no wonder so many robotics teams skip it all together! But parallel-pathing simulation with deployment can mean the difference between a field robotics system that works, and one that does not. This session Polymath Robotics CTO Ilia Baranov will discuss little-known simulation methods that field robotics startups are using for faster, easier deployments, without the usual sim headaches.

Most outdoor robots rely on Global Navigation Satellite Systems (GNSS), benefitting from both the ubiquity and the absolute accuracy of the GNSS. Orientation is another output which can be provided with the use of a dual antenna GNSS systems which can help not only in obtaining heading information but also in reducing the initialization time for having a fast vehicle orientation. However, GNSS, which is built on processing of externally provided signals, is inherently vulnerable to the influence of destroying signals (jamming) or fake signals created to mislead the GNSS receiver. Self-Jamming is also a common problem even with the more professional robotic solutions and this can be caused by self-interference when using other payloads such as lidars, cameras, etc.

This presentation will provide a view on the integration of GNSS with robots and explain how positioning and orientation can be best integrated and used within the ROS2 driver. Attendees will also learn of the different errors that can affect a GNSS receiver, as well as the techniques in the GNSS receiver to overcome those problems. The speaker will also illustrate how a careful choice of the right GNSS receiver can crucially contribute to the robust, accurate and safe performance of an autonomous robot (specially with focus on jamming and spoofing).