The lessons we learn from self-driving will drive our robotics future

Which place does your company rank in the AI adoption curve for AI? To find out, take our AI survey.Robotics is experiencing an explosive growth phase. Robotics is seeing a wide range of applications, from the creative to the mundane. Robots were deployed in new ways to disinfect public spaces, handle infectious materials and provide medical care for patients during the COVID-19 pandemic response.However, the possibilities for new robotics applications are endless. AV (autonomous vehicles) development will accelerate this growth. Why? Why? Because self-driving cars pose the same problem as other types of robots. With its high concentration of capital, talent, and infrastructure, the AV industry is well-positioned to face this challenge.The autonomy challengeAlthough robots are becoming more common, their applications remain limited. One-armed giants were able to perform highly scripted tasks for decades. They were designed for one purpose: adding threads at the pipe's end or spot welding. They weren't flexible enough to handle a wide range of tasks and were not able to work well in unstructured settings. Robots were limited in autonomy and functioned best as an extension of a human actor when deployed in unstructured environments such as surgical settings or aerial drones.However, AVs require great autonomy. There is no human behind the wheel and there are huge stakes. The ability to perceive, plan and act in dynamic, unstructured environments like San Francisco's chaotic streets is essential for AVs. They must be able to communicate with pedestrians, cyclists, drivers and other people, as well as make decisions together.A four-way stop is one of the most common and challenging traffic situations that we all encounter. Despite laws governing how drivers should proceed, most people navigate intersections using nonverbal communication. They look at each other, make eye contact, and nod or wave to one another. An AV cannot communicate with other drivers using these cues. However, it must still be able to decipher their intent and communicate its own. For example, it might slow down and creep forward to indicate its intention to move through the intersection. All the while adhering to traffic laws and making safety-critical decision. This choreography is not possible to plan in advance. Decision-making in AV must be based on human-like social expectations. It should take into account the current and future situations of all actors involved in the scene.The challenge lies in making decisions under uncertainty. This means that you have to make decisions based on imperfect observations and limited knowledge. Robots that are autonomous must observe the world as it is now (imperfect observations), determine how this will change (incomplete knowledge) and then make the right decisions in each situation. This cognitive ability is essential for interpersonal interactions, as human communication requires the ability to understand the motivations and interests of all participants. As humanmachine interactions become more complex, automated systems become more intelligent. We strive to make computers that are communicative as well as decision-making. This is how robots can be transformed from being controlled by humans to being able to collaborate with other machines.Human-robot collaboration: Where it can lead usRobotics has become a more popular industry. As a result, the costs of robotics have dropped, which allows for widespread adoption in a variety of settings. The technology may be familiar in some cases but the application may be new in others. Drones aren't new. However, drones can be used by companies to inspect power lines and collect information to support insurance claims. The same goes for the one-armed giants that are now used as baristas and concierges in hotels instead of spot welders.Automation has greatly benefited commerce. Automation has made materials handling a prime candidate for self-guided vehicles. This is largely due to its high risk sector for humans. Robots outfitted with cameras, lidar and other sensors such as those used to enable AVs perception systems, can navigate factory floors and loading docks safely and efficiently, while avoiding collisions. However, these robots still need to operate in a predictable and structured environment (markers on ground aid them in navigation) and are not responsive. Some have claimed that robots have caused injuries at fulfillment centers in the past few years because they move faster than humans.Robotics has also become commonplace in healthcare settings. Intuitives da Vinci, a robotic-assisted surgical system, is used in 90% of prostatectomies. Robots are becoming more valuable than ever in the operating room, particularly in light of the COVID-19 pandemic. Robots can be used to assist caregivers in lifting patients, performing other tasks, and providing social interaction for the elderly. Robotics are becoming more popular as a STEM educational tool and not just a trendy tech toy. Recent years have seen a lot of interest in the use of emotive robots to treat autism in children.Development of AV is crucialThe global robotics industry is growing rapidly with more players and increased adoption. It's estimated to be worth $100+ billion and will triple by 2021 according to IDC. Driver-assistance technology is a key factor in many of these new vehicles, particularly those at the top end of the market. However, companies working on fully autonomous technology are set to expand the boundaries of robotics in the automotive industry.The possibilities for using these solutions in other robotics applications expand as AV companies face the challenge of human-robot cooperation at the level necessary to bring self-driving cars to market. An AV, like a grandmaster of chess, must consider all possible moves and countermoves for itself as well as other traffic participants. Then it must make safety-critical decisions in an environment that is constantly changing and noisy. It must consider context, such as traffic laws and local norms. Driving in Houston is different from driving in Hong Kong. A successful AV must communicate its goals and intent to people in a natural and intuitive way.The ability to make the right decisions for AVs will allow for complex critical thinking that can be used in other robotic applications. This will enable greater autonomy and collaboration between humans and robots in new and familiar situations. Robots that are able to generate lifelike, engaging behavior autonomously will be safer and more responsive. Both AVs as well as the entire sector will benefit from the shift to humans collaborating with robots rather than them being supervising.Rashed Haq at Cruise is Vice President of Robotics.