Understanding Warehouse Robotics Software
Guest Blog by Dan Gilmore, Chief Marketing Officer of MHI Member Roboteon
While robotic automation continues to be a hot topic in logistics, there are also instances where robotic pilots or proof-of-concept (PoC) initiatives struggle to achieve success.
The same holds true for full-scale deployments, which can take too long to deliver value and ultimately reach the levels of performance expected in the business case.
A theme running across these challenges and more is that well-designed and highly functional software is critical to integrating warehouse robots and operating at maximum effectiveness.
For Autonomous Mobile Robots (AMRs), companies naturally tend to focus on the hardware side of the equation – the physical robot itself, its feeds and speeds, payloads, battery performance and more. This is not surprising: the physical robot is what can be seen, and “speeds and feeds” are easily communicated via spec sheets and videos.
The robot hardware is of course critical to short and long-term project success. But equally if not more important is the software used to plan, integrate, manage, orchestrate, and optimize robotic-enabled distribution processes. Here there are many variables that must be carefully considered at a strategic level, beyond the short-term orientation that characterizes many robotics projects.
Chris Lingamfelter, founder and managing partner of Robot Advisors and recognized robotics expert, recently noted that “Companies consistently underestimate the role and challenges of robotic software despite the fact getting that right is simply essential to a successful project.”
Rapid Advances in Robotic Software
In the last two years, there have been a number of advances in robotics-related software. The most impactful of these are shown in the graphic below:
Important Recent Advances in Warehousing Robotics Software
Individually and together, these advances make the deployment of warehouse robotics faster, easier, and more successful. Designing their warehouse robotics processes and selecting vendors.
Below we highlight a number of specific capabilities for companies to consider when designing their warehouse robotics processes and selecting vendors. Most of these capabilities did not exist until recently in the warehouse robot software market.
They are also increasingly delivered in the Cloud and offered as a platform that includes a broad set of integration and operating capabilities:
Rapid Low Code/No Code Integration: Today companies no longer have to endure the slow, costly, and risky hard-coded approach to integration of robotics with WMS, ERP and other software systems.
The integration between WMS/ERP and the robot software/platform(s) can be enabled through use of AI, which in the right hands can nearly complete the data mappings with only some modest manual work required to finalize the integration, significantly reducing the typical time and effort.
Interoperability: At its core, interoperability involves different types of robots across different vendors operating as if they were all from the same OEM. While there are some initiatives in the robotics sector to develop cross-vendor integration protocols that will make interoperability much easier – and often serve to separate the hardware and software decision – such standards as VDA5050 and Mass Robotics 2.0 are either not widely used or are still under development.
So, what to do in the meantime? Robot vendors typically offer APIs to enable integration to other software platforms. As discussed above, AI-based API mapping technology allows robotics and other warehouse automation systems to be plugged into the platform and thus connect with other hardware and software systems without the need for a major IT integration project.
Pre-Built Execution Applications: Beyond basic integration, robot users need to execute fulfillment and other distribution processes. This typically involves significant code development and is a key factor in the time to deploy and time to achieve value for robotics initiatives.
Some robotics software vendors have developed a portfolio of robot-enabled, configurable execution apps in areas such as piece, case and pallet picking, replenishment and more, significantly reducing the time, cost and resources needed to develop this functionality.
Orchestration Engine: As companies continue to automate distribution centers and deploy materials handling systems of different types and vendors, the need to coordinate and optimize the total flow of goods based on real-time conditions becomes increasingly important. The industry calls this “order orchestration,” and here the goal is to achieve strong capabilities through proprietary algorithms and now AI to maximize the flow of goods in a way that achieves optimal total throughput.
Advanced Analytics: Look for granular, real-time dashboards with performance metrics and throughput data across process areas – or maybe better said “through one pane of glass.”
Simulation and Modeling: Some software vendors leverage simulation tools enable companies to plan for new greenfield or brownfield environments through highly configurable scenario analysis. This simulation will use actual operational or as needed forecast data to estimate picking or other costs from the new robotic automation. The simulation will also calculate expected throughput and many other KPIs by analyzing various combinations of robots and humans, across any time horizon, to help companies plan for success.
Other Opportunities With AI: We of course can’t discuss warehouse robotics in today’s world without a look at the role of AI and ML, beyond using AI to speed integrations, as discussed above.
Let’s start by saying this: we are still early in the ball game. What’s more, in many cases, traditional software smarts – as delivered by advanced heuristics and algorithms – have set a high bar in terms of operational efficiency.
That said, areas where AI/ML look promising include:
Forecasting and Pattern Recognition: Involves predicting future order patterns based on historical data and current trends.
Task Scheduling and Release: Workload balancing in the warehouse across robots, people, automation based on real-time conditions, factors such as customer, carrier cut-off times, etc. – the orchestrion capabilities discussed above.
Dynamic Path Optimization: Adjust travel paths and picking routes for robots and operators based on real-time operational data from the floor as well as order profile, priorities and other factors.
Order Clustering: Group multiple orders together for picking and overall throughput efficiencies, considering multiple parameters, including order types, SKU, storage locations, customer, etc.
There are other opportunities. For now it’s probably just best to evaluate where a given vendor sees their AI/ML developments going.
Also note Gartner has come up with a term that defines this type of robotics software: Multiagent Orchestration Platform – I recommend adding it to your lexicon.
We have entered the warehouse robotics era. It will be powered by software.
