Why are autonomous transport robots a real game changer in logistics?

To many logistics experts it might seem like old wine in new skins: Everybody is talking about transport robots in logistics nowadays. But haven't AGVs been around in logistics for decades? And yes. They are quite right to be honest. Frankly spoken since 1953 when Arthur Barrett presented his “Guide-O-Matic” wich is considered as the very first AGV.

But thats exactly the important notion here: AUTOMATED Guided Vehicle or AGV.

This means that the transport vehicles need some sort of external guidance device like wires or magnets in the ground, lines to follow or even laser markers on factory pillars to get some orientation.

Now there is a quite new class of transport vehicles arising, called AMRs or AUTONOMOUS Mobile Robots.

In our definition a robot is autonomous when it can maneuver from point A to point B, bypass obstacles and navigate without any additional installation in the building.

As one example you can see our Smart Transport Robot - or short STR - in the video that I posted last week: STR - Smart Transport Robot

To achieve this task, the STR has to fullfill several prerequisites:

1. The STR has to localize itself in the factory.
2. The STR needs its own onboard map of the surrounding.
3. The STR manages to do its own onboard path-planning.
4. The STR needs to exercise all safety relevant calculations.
5. The STR needs to run the neural networks to detect the load carrier.

Localization is one of the major key points when it comes to autonomous transports and is mostly achieved through a new type of algorithm called SLAM (Simultaneous Localization And Mapping). Basically what the algorithm does, is that it uses the safety scanner to detect its surrounding and compares it to the saved map of the building, and therefore calculating backwards the position of the vehicle. Thats sounds quiet simple but there are some pitfalls like very similar building structures where the algorithm just finds too many good fits in the map for one position and gets confused. Or on the other hand when you have surroundings that change during the day. This could be a block storage where sometimes containers are stored and sometimes its empty (wich is of course quite usual in logistics buildings) or high traffic of other vehicles like forklifts and tugger trains. In our own vehicle we were facing an additional task: Because we wanted to start with a base version of the STR (to keep the cost at a minimum) we needed to navigate with only one safety scanner in the front of the robot. The safety bumper in the back does obviously not generate any additional data that we can use for localization purpose. Therefore we used a very sophisticated SLAM module from our Fraunhofer IPA development partner. The IPA SLAM also manages to sync the path-planning of the robot to the different safety zones of the scanner (e.g. loaded or empty STR). It also takes into account the variable enviroment (Long-Term-SLAM) and enables the STR to bypass obstacles in a very dynamic manner (Elastic-Band). As the module is based on ROS (Robot Operating System) we are gaining a high flexibility regarding future adaptions and interfaces.

So what are the big benefits of such a autonomous transport robot like the STR?

We actually believe that autonomy is one key enabler for managing large and robust fleets of robots. The reason for that is the resilience of the robots itself. Even if the STR loses the WiFi connection (wich can happen in large plants or even in smaller ones at the transition from one router to the next one) it does not stop and can fulfill its task to the end. Another key aspect of autonomy is the possibility to react to an obstacle on the path, and this happens all the time in logistics operations. That doesn't mean that the robots are driving around like crazy in our plants, there are certain rules (e.g. restricted areas, one or two-way roads). But the robot is allowed to go on the other side of the road if the road is blocked, and we learned that this is a huge gain when it comes to system robustness.

Also very valuable for logistics operations: Because of its object recognition capabilities the STR can autonomously pick-up and drop its load directly at the point of use without any additional handling and without an exact positioning of the roller carrier.

But maybe the biggest advantage that we get from autonomy: It enables the STR to receive its orders and informations from a cloud system - we call it: Services ATS (Autonomous Transport Systems). In our opinion this is a major leap in flexibility, cost efficiency and future viability.