From guiding to self-driving

According to Statista, there were over 187,000 automated guided vehicles (AGVs) around the world during 2022. While these driverless vehicles might be in the distant future for consumers, they ensure manufacturing plants can improve productivity, manufacturing flexibility and production flow.

AGVs are mobile robots that follow a set of rules, be it signals or markers, during navigation. The first AGV was introduced in the 1950s, by Barrett Electronics, and was guided by a wire in the floor that performed a simple towing action. The market has rapidly increased during the 21st century, and these mobile robots are now common across several industry sectors.

Why AGVs?

AGVs are capable of reliably delivering necessary materials without the need for human assistance, ensuring that there is minimal interruption to the production line. However, self-guided vehicles don't just transport materials, they are also used for work-in-process applications and transporting the completed goods.

AGVs are also widely used to support the process of picking and handling — both inbound and outbound — for replenishment. For example, here AGVs might be used for transporting inventory from the receiving bay to the storage locations. They could also transport materials from long-term storage locations to a forward-picking location for the replenishment of stock. This ensures that sufficient inventory will be readily available to the pickers, which increases the efficiency of the order-picking process. 

Depending on the application and requirements, AGVs can function in isolation or in fleets. This makes the use of AGVs scalable according to need, so a plant manager can make specific decisions on the number of vehicles in a facility. Furthermore, when equipped with sensors for traceability, AGVs allow plant managers to monitor the position of each individual vehicle and therefore track the movement of materials around a facility.

Pick-up, transit and delivery of items can be time stamped during this process to further enhance tracking abilities. This data can then be incorporated into the company’s enterprise resource planning (ERP), or materials resource planning (MRP) systems. However, there are several types of vehicle navigation methods depending on the application. A plant manager can therefore select a very simple system, like the earliest AGVs, or opt for a more advanced navigation method. 

Guided navigation

The original AGVs moved around the factory floor by a wired method. Here, a wire is embedded into a slot in the floor and transmits a radio signal, which is detected by a sensor on the AGV. Quite simply, the AGV is guided around the facility following the wire. This navigation technique is in use today, although there are now more sophisticated methods for plant managers to go for. 

For example, some AGVs use guide tape, which is either magnetic or coloured, to traverse their environment. Sensors on the AGVs detect the tape to guide the vehicle. Guide tape uses also include laser target navigation, where reflective tape is fitted on walls, poles or machines so the AGV can calculate distance using a laser transmitter and receiver. This is a significant benefit compared to the wired method, as it is easier to change the vehicle’s route as the process of rearranging the tape is simpler. 

Inertial navigation is another navigation method. This approach uses reference points that are embedded in the factory floor as x,y coordinates, ensuring the AGV can use information from a sensor, a gyroscope and a wheel encoder to determine location. Like the previous method, changes can be made to the pathway by merely altering the reference points, providing far more flexibility. Consequently, this does require some factory infrastructure changes as the vehicle cannot make independent route-planning decisions. 

Self-driving 

The next step is open path navigation. This allows the vehicle to move freely from one place to another, shifting from a guided vehicle to a self-driving vehicle. Even despite infrastructure changes, traditional AGVs that perform defined, pre-programmed movements around a facility make it difficult to change the vehicle’s route. More flexible and intelligent vehicles were introduced as a result, which ensure decisions can be made in situations that haven’t been faced before. 

In an ever changing environment, a self-driving vehicle is simply a better fit. These AGVs operate independently from a driver or a fixed pre-programmed input, instead using laser-based perception and navigation algorithms. These help the AGV to dynamically travel around a factory, with far greater flexibility. 

In addition, the likelihood of the AGV making errors can be reduced by integrating an on-board programmable logic controller (PLC). By linking to the central control system, the AGV can analyze the reliability and efficiency of its routes and adapt them accordingly. What’s more, the AGV can use machine learning algorithms to be more efficient when encountering new situations. 

An example of a self-driving AGV is Clearpath Robotics’ OTTO, a self-driving vehicle that can handle 1500 kg of cargo, plus another 400kg of cargo, for a total of 1900 kg of cargo. OTTO can also adapt to take the best route, avoiding collisions as it moves. 

These vehicles can also use vision-based guidance systems, using cameras to act as eyes, for plant managers to gain a 3D virtual view of the environment the equipment is operating in. However, for plant managers to be aware when an AGV comes across anything unplanned or unusual, sensors are a critical component. 

Equipment breakdown is inevitable, though. This is why developing a relationship with a reliable parts supplier can ensure the technology essential to the production process is continually in operation, without the inconvenience of unplanned downtime. 

AGV technology has come a long way since Barrett Electronics in the 1950s and plant managers have access to a range of different types of AGV to suit different needs and different budgets. Modern self-driving AGVs are now becoming increasingly commonplace in manufacturing facilities and are helping plant managers improve throughput and efficiency in ways that were unimaginable a couple of decades ago.