Giving robots the ability to work together in a dynamic environment could be game-changing.
Ants are stupid! I don’t mean to be mean; that is just a fact of nature. Individual ants are some of the least intelligent creatures in the animal kingdom. For example, two ants can be fighting over the same piece of food, which they both mean to take back to their anthill. If the two ants disagree on the path back home, they will keep fighting over that piece of food until they both die… of starvation… while holding food in their mouth. That is a special kind of stupid.
However, an ant colony is unbelievably clever. The ants are capable of building enormous, intricate structures that rival human cities. They grow and farm their own food; they have a well-defined social system, and even grow a type of insect named an aphid which they milk!
So why am I talking about ants on a blog dedicated to robotics? Because I believe that robotics fleets can take several “stupid” robots and turn them into a colony of hyper-intelligent bots.
Think of a retail store that employs several robots - one to greet the customers at the entrance, five robots who “work the floor” by answering people’s questions and directing them to the product they need, another robot at the register, and several smaller ones to take stock and refill the shelves as needed. All in all, you would say that this store is running at peak efficiency. Every robot does its job, and all is fine.
However, a robotic fleet can take the whole operation to another level. While the customer is consulting the “floor” robot, The information about the conversation is uploaded to the FMS (Fleet Management System) that sends it to every other robot in the fleet. Once the human decided on a specific item and took it off the shelf, the registry robot has already entered the item into the final tab of this particular customer, and one of the stock robots has already grabbed a new replacement item from the storage room and is on his way to put it in its place.
So while the customer is taking items off the various shelves, their invoice is already being printed, replacement items are being restocked, and the FMS is trying to anticipate what other items this particular customer might be interested in.
Let’s say that the customer is done shopping and is ready to pay. However, someone else is still in the process of paying at the registry. In this case, all the information about the various items is sent directly to the “floor” robot, which can now behave like a registry robot. And while the customer pays, one of the “stock” robots quickly runs up to the registry to grab the invoice and take it to the customer.
A well-organized fleet of robots can go beyond their roles and programming and start improvising solutions to unpredicted problems.
