This is the first post in a blog series on the challenges that factories, warehouses, and logistic centers face when deploying autonomous mobile robots (AMR) and a possible solution for these challenges.

AMR Challenges in Price

We’ll start with some facts: manual material handling represents 20–25% of total factory workforce, so increasing productivity is a prime objective of every manufacturing company or logistic center. COVID-19 has reinforced this need, boosting demand for mobile robots in industry, logistics, and warehouses.

At the same time, sensing, visual and computational technologies have matured dramatically; AMR producers are proliferating and factory floors are incorporating advanced AI and IoT platforms. It seems as if Industry 4.0 is here and everything is ready for the AMR revolution.

Yet when we look at the numbers, only a few thousand AMRs are sold worldwide each year. ABI Research estimates that out of a million forklifts sold annually, only 2,000 are automated (a rate of 0.2%). Gartner puts adoption rates at 1–5% for the whole AMR market, but when you take Amazon out of the equation — the numbers decrease dramatically.

Why aren’t AMRs catching on? In this post, we will tackle one major reason — price.

Expensive Machines

AMRs are expensive machines (although admittedly less so than their predecessors — AGVs).

To roam freely, AMRs require expensive equipment: sensors (visual, speed, sound, motion, light), a navigation system to read and interpret the situation (so they can change their trajectory and speed) and a management system to oversee and synchronize the AMRs on the floor.

This brings the cost of a unit to anywhere from $25,000 to $0.5M (depending in large part on the application and functionality). A simple high load AMR with a payload of 100–500 KG costs about $30,000 per unit. To this add accessories, feature options, software (and software updates), training, and services.

As with most hardware equipment, AMRs do not enjoy economies of scale. On the contrary: as you buy more, operating costs go up because you need to hire more overseers and engineers and install larger and more sophisticated control centers (see the video linked below).

ABI Research sites lack of scalability as a major reason why AMR vendors are not profitable. The AMR market is also highly fragmented, withover 280 different manufacturers targeting specific verticals. Lack of standardization has a price — literally — in higher costs per unit.

Expensive Deployment and Maintenance

AMR deployment requires changes to the floor layout, to create clear paths and accessible workstations and to “cage off” areas AMRs cannot and should not enter. There is the cost of human resources: whoever will operate and manage the AMR fleet will need to learn how to use the software, map the robot routes, oversee recharging and maintenance, and continually supervise the vehicles.

Integrating and tweaking the management systems takes time. You don’t just come in on Monday and the AMRs are speeding across the floor. You need to map the floor, routes and stops, where you recharge, when you recharge, maintenance schedules (those expensive cameras need to stay clean), employee training — the list goes on.

Then there is the trial phase: you start with one AMR, then add another, then another. Then you check: do the workers utilize it? Are they adhering to safety protocols? Is the load optimal? Often, it take weeks for AMRs to start operating, as the team solves localized problems. All this tweaking is in fact the main reason for deployment delays.

Take a look at this interesting video by Seat, which displays impressive integration of mobile robots into a car manufacturing plant. Incidentally, the video also shows the large situation room and the staff continuously supervising the robots.

Seat produced the video to showcase their major savings (a 25% reduction in operating costs). However, the video also displays the considerable initial investment required to deploy an AMR fleet. AMRs are currently the privilege of large corporations with deep pockets and administrative capabilities, in short — the Amazons and Alibabas of this world. SMEs for now are left out of the AMR revolution.

Unforeseen Expenses

Every large-scale project has unforeseen expenses, and in the case of AMRs these are mostly recalls and robot-human relationship problems.

The more complex a machine, the higher are the chances of failure, and AMRs are no exception. Recalls are common and even if you’re insured, replacing a machine means extra work and lost productivity. Ongoing maintenance eats into OPEX, and according to ABI Research are a major reason for the current reluctance to automate. If CAPEX costs a fortune and OPEX doesn’t improve, then what’s the point?

In addition, workers and especially production workers need to learn to live with robots and to utilize them — no small feat. If workers don’t feel comfortable and safe, and know that the AMR is there to help them and not to replace them, the company will not experience productivity growth or cost reduction.

The Solution

The current AMR paradigm requires expensive navigation systems and sensors on each robot. As long as this paradigm continues, mobile robots will continue to be prohibitively expensive and challenging to operate.

MusashiAI is working to create a different paradigm whereby the “eyes” and “ears” of the AMR move to the facility ceiling, and its navigation and management “brain” to a sophisticated, AI-based, central control-tower system.

The MusashiAI central system continuously watches and navigates the autonomous mobile robots on the floor; optimizing paths, detecting hazards, and managing multiple vehicle types (such as cleaning and transporting).

Freeing AMRs from expensive embedded hardware leads to a dramatically lower cost-per-unit. Our technology is platform agnostic and can control any type of electrical vehicle. Deployment becomes easier, faster, and cheaper. The cost of failure goes down and adoption rates go up.

We at MusashiAI believe that our central navigation and fleet management system is the next stage in the Autonomous Mobile Vehicle evolution; and that our systems will enable many factories, logistic centers and warehouses to enjoy the benefits of the AMR revolution — even if they are not Amazon or Alibaba.

Contact us to learn more before planning your next AMR fleet.

Maestro@634.AI