Truly Smart Robots Know When To Ask For Help
Vecna Robotics develops autonomous mobile robots (AMRs) used in distribution, logistics, warehousing and manufacturing (the company first cut its teeth in hospitals and other health care facilities). Their secret sauce is the software; after all, explains founder and Chief Innovation Officer Daniel Theobald, robotics is 90-95% software. Vecna Robotics implants its software brains in existing equipment like forklifts and pallet trucks, making them autonomous – though the company does build some hardware of its own.
What sets Vecna Robotics apart is the combination of an ambitious vision with a pragmatic attitude, articulated in three key principles:
- First, “the key to success with robotics lies in focusing its application to the areas where it can yield the biggest benefits now”, says Theobald. We need to understand what robots can and can’t do today in a work environment and deploy them where they can give the greatest immediate efficiency gains. In fact, Theobald thinks a key reason why the adoption of robots has not moved faster is that too often we try things that robots are simply not yet ready for.
- Second, we need to think in terms of robots and humans working together. Not only because attempts at full automation have so far failed; but more importantly because humans and robots have different abilities, and to maximize productivity and profits we should optimize the combination of both sets of skills, rather than thinking of robots as a substitute for workers.
- Third, successful adoption of robotics requires changes in operations and management practices. As we understand what robots can do, and how they can best collaborate with human workers, we need to reorganize operations accordingly.
While Theobald says we should be mindful of what robots cannot do, Vecna Robotics’s AMRs have some impressive capabilities.
They are truly autonomous, in that they do not simply follow a pre-programmed routine, but are able to react and adapt to unexpected circumstances to get the job done. Vecna Robotics’s software allows AMRs to decompose their “mission” (the goal to accomplish, such as retrieving a pallet from location A and moving it location B) into actions and tasks; as they move around the work environment, the robots have the awareness and autonomy to adapt to unplanned situations, for example navigating around an obstacle, or searching a wider area if an item is not in its expected location. While they do so, they monitor whether they are still on track and on time for completing their mission.
They know when to ask for help: when they face a difficulty they cannot solve by themselves, they call on a human supervisor to step in. Not literally, because the supervisor might be in a completely different location. In fact, Theobald points out that their technology allows them to resolve most situations without the need for local intervention. The human colleague can see the robot’s situation through a set of cameras, replay the robot’s most recent actions and see what attempts the robot has already made to resolve the situation; the supervisor can then instruct the robot on how to proceed.
The robots learn on the job and from each other: as they go about their work they accumulate more knowledge about their environment, and Vecna Robotics’s multi-agent AI-based engine, named “Pivotal”, allows them to share this knowledge with their robot colleagues.
They are versatile: Theobald notes that the same robots can be rapidly reconfigured to operate in different environments. You can use them to complete a set of operations in a warehouse in the morning, then reprogram them and deploy them in a distribution center in the afternoon for different purposes. This versatility is proving particularly attractive as companies try to cope with the latest wave of disruptions by investing in flexible automation solutions to make their operations more resilient and adaptable.
Some of the costliest supply chain disruptions, however, are happening outside a company’s own warehouse; the massive delays in major ports, for example. Could robots help there as well? Daniel Theobald thinks it’s only a matter of time. Future application of robotics, he says, will follow what he calls the “rule of the three Ms”: everything that Manufactures, Moves or needs Maintenance will eventually be automated. The best way to move forward, says Theobald, is by applying robotics to functions that can yield massive amount of data about their real-world use, since Artificial Intelligence still largely boils down to pattern recognition.
Daniel Theobald is passionate about robotics; he is a cofounder of Mass Robotics, a Boston-based non-profit committed to fueling the development of robotics. At the same time, he insists that “everything we do needs to be about people.” There is a genuine commitment behind this: since inception, Vecna Robotics has paid its employees to devote 10% of their time to community service. But there is also an unshakable confidence in people, a conviction that no matter how fast robotics advances, humans will always be needed, and that robotics will make people better off.
As today’s economy is caught between the dystopian fears of mass automation and the current reality of worker shortages, Vecna Robotics’s pragmatic focus on human-robot collaboration shows the best way forward.
Post a Comment