Yes, our ANT technology includes a safe and highly configurable form of obstacle avoidance called SmartPass. Learn more about SmartPass, and explore when obstacle avoidance makes sense, with this detailed guide.
Obstacle avoidance is a growing requirement in mobile robot tenders and RFQs. But what does it really mean in practice and is it always the most efficient option?
In this guide we explain:
Automated vehicles such as AGVs and AMRs typically navigate in one of two key ways:
1. Path following
Vehicles follow a path; either physical (as with magnetic tape), or virtual (as in the case of ANT navigation).
2. Obstacle avoidance
Vehicles determine their own routes, dynamically, in real-time.
Generally speaking, AGVs follow a pre-defined path of some sort, while AMRs continually sense the environment in order to find their own route independently.
The proof of this difference? The ANSI/RIA R15.08-1-2020 standard, which defines the safety requirements for the design and integration of industrial mobile robots, states:
“The fundamental difference between AGVs and AMRs is characterized by how they traverse the specified operating environment. An AGV traverses the specified operating environment automatically along predefined guide paths (virtual or physical) using collision avoidance”, whereas the standard defines AMRs as able to “traverse the specified operating environment by detecting obstacles using sensors and adjusting paths by computing an obstacle-free path through free space rather than using a predefined path.”
Let's quickly compare how AGV-style path following and AMR-style obstacle avoidance work, side by side:
Important to know: 'ANT driven' vehicles can use either virtual path following or our version of obstacle avoidance (SmartPass), even in the same installation.
Now we've covered how these technologies differ, let's explore their pros and cons:
A useful way to compare these different navigational approaches is to look at the ideal use cases for each.
Pure obstacle avoidance—whereby a vehicle continually creates its own route—can make sense if a robot is small and is used in a busy area.
A good example would be a cleaning robot, such as those by Cleanfix. These robots typically need to cover (clean) a set space, but the order in which they cover it is not important.
In environments like commercial centers and airports, the probability of Cleanfix's robots encountering an obstacle, particularly a person, is high. Therefore, if the robot stopped at every obstacle (AGV-style), the floor would never get scrubbed. In this case, using pure obstacle avoidance is a great fit.
Path following makes more sense in industrial settings, such as material handling use cases in warehouses and production plants.
The key to success with path following? Having well trained staff on-site who can hear, notice and respond to blocked vehicles by moving blockages immediately (or ideally not creating them in the first place).
In the case of large vehicles carrying heavy payloads, path following is a great fit, since staff are often less anxious when vehicle behavior is predictable and clearly defined.
Path following also suits space-limited sites such as brownfield factories, where moving around obstacles can quickly lead to vehicles getting stuck (deadlocked).
Despite the real-world observations above, the belief that obstacle avoidance is the most efficient approach is common.
Customers see AMRs in highly polished marketing videos doing their thing, ‘swarming’ intelligently around, and they intuitively believe this approach must be more efficient.
But in most cases this isn't true. Despite its appeal, obstacle avoidance often underperforms in industrial environments.
The main issues with AMR-style behavior in industrial environments are:
Vehicles (AMRs) that rely on obstacle avoidance as their primary mode of navigation typically move more slowly than AGVs following their pre-defined paths.
AMRs move slowly because they are continually sensing and calculating where to go. They have to then reduce their speed to react to obstacles (which they meet more frequently, since they go wherever they like), and they must take longer routes as they bypass them.
The efficiency of fleets of obstacle avoidance vehicles is typically low, due to the absence of coordinated traffic control.
AMRs are built, from the ground up, with obstacle avoidance as their key navigational behavior, rather than having traffic management integrated onboard.
In other words, they operate independently, rather than having a fleet manager inform them of blockages and manage their movements and priorities at intersections.
The result of the lack of strong traffic management is traffic deadlocks. AMRs often become blocked by each other, stopping transport flows.
Some AMR suppliers try and mitigate the risk of deadlocks by layering low-level traffic management on top of obstacle avoidance—for example by adding basic rules like preferred direction areas and exclusion zones—but these measures only deliver marginal gains at best.
At BlueBotics, we have seen hundreds of end customers decide between path following and pure obstacle avoidance, and we've also seen the consequences of these decisions.
The result: in most cases, virtual path following based on strong traffic management is far more efficient than pure obstacle avoidance.
In one example, a North American tire manufacturer operated a fleet of 37 mobile robots, all of which ran using pure obstacle avoidance (whose navigation technology was supplied by another company, not BlueBotics). The end user was displeased with the overall performance of the system, citing common deadlocks and general inefficiency.
These robots were then retrofitted with ANT navigation, and their core navigational approach was changed from obstacle avoidance to solid, pre-programmed virtual path following.
Result: The same fleet achieved higher throughput with fewer vehicles, simply by switching navigation approach.
The tire producer required seven fewer AGVs, and even with that reduction the system is 10% more productive than previously.
To be clear, these were the same vehicles. The only difference was between them using virtual path following in place of obstacle avoidance.
While obstacle avoidance is clearly not the optimal approach for industrial applications, that does not mean that path following works perfectly by default. For example:
In these cases, the ideal scenario would be for the path-following robot to move smartly around a blockage before returning immediately to its usual path. The best of both worlds in other words.
With ANT inside your vehicle, you can achieve exactly this type of combined operation using a feature called SmartPass.
SmartPass is an advanced feature within our ANT software suite that adds controlled obstacle avoidance to virtual path following.
In short, it combines:
Unlike AMR-style obstacle avoidance, SmartPass ensures that all avoidance maneuvers remain controlled, predictable, and aligned with overall fleet efficiency.
This video explains all:
Rather than layering basic traffic management over obstacle avoidance functionality—which AMR producers have attempted with limited results—SmartPass does the opposite: it adds smart, configurable obstacle avoidance to ANT navigation’s default ‘virtual path follower’ mode.
This means the powerful traffic management functionality of our ANT server fleet manager is also applied to SmartPass obstacle avoidance maneuvers.
There are three key benefits of this approach:
1. Efficiency-focused movement
2. Minimizes deadlocks
3. Fully configurable
Safe and prudent, SmartPass is fully configurable, to suit every customer and site. Your team can define, for example:
In most cases, at BlueBotics we recommend choosing pure path following. However, the following framework can also help you make, or confirm, your approach.
Scenario 1. If...
Choose virtual path following.
Scenario 2: If…
In addition to the requirements above, there is also a reasonable chance of objects blocking the progress of your robots…
Choose virtual path following + SmartPass
Scenario 3. If…
Vehicles need to cover an entire space and the order in which they achieve this is not important (e.g., as in the case of cleaning robots)…
Choose pure obstacle avoidance.
Reminder: All of the navigation modalities above are available within our ANT lab software's configuration settings, with ‘virtual path following’ the default choice.
Our BlueBotics engineers are happy to discuss this topic with you in more detail. Just get in touch here to set up a call.