Latest / Mechanical Engineering Made Simple / (#114) Fixing a Material Handling Disaster

Transcript

(#114) Fixing a Material Handling Disaster · Mechanical Engineering Made Simple

• 0:00
  Welcome back to mechanical engineering made simple.
• 0:02
  Glad to be back. So today we are heading out on a
• 0:05
  really unique deep dive mission. We are.
• 0:09
  We're going to explore the invisible choreography that
• 0:12
  basically keeps the entire modern world running.
• 0:16
  I'm talking about the massive, unseen systems of material
• 0:20
  handling. Oh yeah, the silent partner of
• 0:22
  manufacturing. Exactly.
• 0:23
  It's the lifeblood of literally every physical product you own,
• 0:27
  yet you know it's a discipline that rarely gets the spotlight.
• 0:30
  At least not until the system completely breaks down and
• 0:33
  grinds production to an absolute halt.
• 0:36
  Which happens more often than you'd think.
• 0:38
  And to truly understand the engineering rigor behind it, we
• 0:41
  aren't just going to read off a list of textbook definitions.
• 0:46
  Instead, we're taking you on a story driven walkthrough.
• 0:49
  We're going to drop you right into the middle of a common real
• 0:51
  world engineering disaster. I love this approach.
• 0:54
  Yeah, picture this. You've got a sprawling, high
• 0:56
  mixed discreet manufacturing plant and this facility is
• 0:59
  currently suffering from plummeting overall equipment
• 1:02
  effectiveness. I mean massive delays, damaged
• 1:05
  goods, sky high operational costs.
• 1:09
  A total nightmare scenario. Absolutely.
• 1:11
  So our mission on this deep dive?
• 1:13
  To use core material handling principles to diagnose the chaos
• 1:17
  on that floor and systematically engineer a fix.
• 1:20
  So imagine you are standing on the floor of this factory right
• 1:24
  now. It is.
• 1:26
  It's an absolute crisis. Total chaos.
• 1:28
  You've got WIP work in process inventory just piling up in
• 1:32
  these random staging areas. Forklifts are completely
• 1:35
  deadlocked in the main aisles. Horns honking, people yelling,
• 1:38
  and. Then you look over and you've
• 1:39
  got these highly skilled CNC machinists leaving their actual
• 1:43
  stations wasting valuable time manually carrying like 50 LB
• 1:48
  bins of raw materials clear across the facility.
• 1:50
  It hurts to even picture it. It's.
• 1:52
  Loud. It's chaotic and management is
• 1:53
  pulling their hair out because lead times are slipping, but
• 1:55
  nobody seems to actually know where the real bottleneck is.
• 1:58
  And here is where we see the classic rookie mistake.
• 2:01
  The knee jerk reaction. Exactly.
• 2:04
  The common engineering reflex in a panic situation like this,
• 2:07
  especially, you know, when management is breathing down
• 2:10
  your neck, is to throw capital expenditure at what is
• 2:13
  fundamentally a process problem. Right.
• 2:15
  They think, well, we have a movement issue, let's just buy a
• 2:17
  bunch of new equipment, yes. So they go out and procure a
• 2:21
  fleet of brand new forklifts. Which is the worst thing you
• 2:24
  could do. It really is.
• 2:26
  As any seasoned industrial engineer will tell you, buying
• 2:29
  equipment before fully mapping the system dynamics only treats
• 2:33
  the symptoms. You are essentially just adding
• 2:36
  more vehicles to a traffic jam. You're just accelerating the
• 2:39
  chaos. Precisely.
• 2:41
  So as the engineer brought in to fix this mess, your proper first
• 2:45
  step is identifying and defining the true scope of the problem.
• 2:48
  You can't just look at the material handling equipment in
• 2:50
  isolation. You have to look at the whole
• 2:52
  picture. Right, you have to understand
• 2:53
  the products themselves, the intricate manufacturing
• 2:56
  processes, the specific TAC times and production schedules.
• 2:59
  And the actual 3 dimensional layout of the facility.
• 3:02
  Yes, you have to look at the factory as a living, breathing
• 3:06
  Organism. That's a great way to put it,
• 3:08
  and you can quickly categorize the operational failures on that
• 3:11
  floor if you break them down into the five fundamental
• 3:14
  aspects of material handling. Let's run through those.
• 3:17
  Every failure out there is a violation of at least one of
• 3:20
  these. First up, there's motion.
• 3:22
  OK, this isn't just about moving things, it's about the
• 3:25
  kinematics of the floor, moving things efficiently, safely, and
• 3:29
  with minimal energy. Right now, having people
• 3:32
  manually carry heavy boxes is a massive ergonomic liability.
• 3:37
  And a terrible use of kinetic energy.
• 3:38
  Absolutely. Then there's time.
• 3:40
  This is your just in time delivery aspect.
• 3:42
  You need materials arriving at the workstation exactly when
• 3:46
  they are needed, not. Really not late.
• 3:47
  Right. Not 10 minutes early, creating a
• 3:50
  localized bottleneck that clogs up the operator's workspace.
• 3:53
  And definitely not 10 minutes late leaving $1,000,000 machine
• 3:57
  sitting idle, bleeding money. The third aspect is place.
• 4:00
  Materials must be positioned correctly for use.
• 4:03
  Point of use presentation. Exactly.
• 4:06
  If an assembly worker has to turn around, walk 20 feet, and
• 4:09
  bend over to grab a component from a pallet on the floor for
• 4:12
  every single cycle, that is a severe failure of place.
• 4:16
  It adds up so fast. It does. 4th is quantity.
• 4:20
  You have to deliver the correct volume to match the process
• 4:22
  demands, aligning your batch sizes with your single piece
• 4:25
  flow goals. And finally, space.
• 4:28
  This is about cubic utilization, not just square footage.
• 4:31
  Oh, this is a huge one. In our disaster factory, boxes
• 4:34
  and bins are spread out all over the floor in a single layer.
• 4:37
  They're paying for a 30 foot ceiling, but only using the
• 4:40
  bottom 4 feet. It's a complete waste of the
• 4:43
  facility's vertical volume. Now that we frame the scope of
• 4:46
  the disaster, we need to transition into diagnosing the
• 4:49
  root causes. We know the system is broken,
• 4:52
  but we need hard data to prove exactly where and why.
• 4:55
  This is where we put on our lead investigator hats.
• 4:57
  We have to systematically collect data on this chaotic
• 5:01
  factory floor. You absolutely cannot just guess
• 5:03
  based on an afternoon of walking around with a clipboard.
• 5:06
  No, you need systematic charting to capture the true flow
• 5:09
  dynamics. Let's talk about the specific
• 5:11
  diagnostic tools used to uncover these root causes.
• 5:14
  The foundational tool here is the Flow process chart.
• 5:18
  This involves tracking a specific component or sub
• 5:20
  assembly and tabulating every single step it takes using
• 5:24
  standard American Society of Mechanical Engineers or ASME
• 5:28
  symbols. Right, the ASME symbol.
• 5:30
  Yes, those symbols categorize every moment into operation,
• 5:34
  transportation, inspection, delay, and storage.
• 5:37
  And when you actually chart this out for our broken factory, the
• 5:40
  data usually reveals something totally shocking.
• 5:42
  It's always eye. Opening you might track a widget
• 5:44
  and find it undergoes, say, 4 value add operations, but it
• 5:48
  requires 12 transportations. And a massive percentage of its
• 5:52
  life cycle falls under that ASME delay symbol.
• 5:54
  Exactly. The actual value add time might
• 5:57
  be 5 minutes, but it takes three days to get through the plant.
• 5:59
  It's just sitting in queues waiting.
• 6:01
  To visualize that inefficiency, we take that data and build a
• 6:05
  flow diagram. We superimpose those exact ASME
• 6:08
  symbols directly onto the floor plan of the facility.
• 6:11
  So the abstract data becomes a tangible map.
• 6:14
  Suddenly you can physically see the chaotic spaghetti string
• 6:17
  paths the materials are making crossing back and forth over
• 6:20
  each other, creating massive cross hauling waste.
• 6:23
  But the flow diagram only shows us the current state.
• 6:27
  To figure out how to rearrange the plant, I'd bring in a from
• 6:30
  to Chart right? It's essentially a cross matrix
• 6:33
  that tracks the required number of physical moves or the flow
• 6:36
  intensity between different departments.
• 6:38
  Yeah, if the matrix shows that the milling department sends 500
• 6:42
  loads a day to the grinding department, those two nodes need
• 6:46
  to be physically adjacent. True, but adjacency isn't solely
• 6:50
  dictated by flow volume, which is why we pair the from to chart
• 6:54
  with an activity relationship chart.
• 6:56
  Right, because of the qualitative constraints.
• 6:58
  Exactly. You might have a high flow
• 7:00
  volume between a heavy stamping press and a delicate coordinate
• 7:03
  measuring machine or CMM. Oh, you do not want those next
• 7:07
  to each other you. Absolutely cannot put a
• 7:09
  vibrating stamping press next to sensitive metrology equipment.
• 7:13
  The Activity Relationship chart maps those proximity
• 7:15
  restrictions. All right, the data is in and
• 7:18
  the bottlenecks are glaringly obvious.
• 7:21
  The factory is violating several of the core principles of the
• 7:24
  Material Handling Institute. It's a textbook disaster.
• 7:27
  They're moving single items haphazardly, wasting overhead
• 7:30
  space, and honestly ignoring basic physics.
• 7:34
  So how do we actually start untangling this mess on the
• 7:36
  floor? We systematically apply the core
• 7:39
  principles to engineer a new state.
• 7:41
  The first major intervention targets the floor clutter and
• 7:44
  the sheer number of trips being made.
• 7:46
  We apply the unit size principle.
• 7:48
  Unitizing the chaos. Right, instead of having
• 7:50
  operators handle hundreds of individual cartoons, we engineer
• 7:54
  unit loads. And building a proper unit load
• 7:57
  is a science in itself. It's not just piling boxes on a
• 8:00
  standard wooden pallet and hoping for the best.
• 8:03
  Definitely not. You have to account for load
• 8:04
  stability and the center of gravity.
• 8:07
  If you stack cartoons in a standard block pattern where one
• 8:10
  column of boxes sits perfectly flushed on top of another, they
• 8:13
  lack lateral stability. They will tip over the second a
• 8:16
  forklift takes a corner too fast.
• 8:18
  That's why we specify interlocking pinwheel patterns.
• 8:21
  The overlapping layers create structural integrity, much like
• 8:24
  laying bricks. Makes total sense.
• 8:26
  But even an interlocked load needs to be secured to survive
• 8:29
  transit dynamics. The entire unit load must be
• 8:32
  consolidated using automated stretch wrapping which applies
• 8:36
  pre stretched film to create a holding force or high tension
• 8:39
  banding. And once it's unitized, a single
• 8:42
  industrial truck can move what used to take an operator 50
• 8:47
  individual trips. Exactly.
• 8:48
  Now that we've consolidated the loads, let's look at the packing
• 8:51
  area. Currently they have powered flat
• 8:54
  belts moving boxes a short distance to the loading dock.
• 8:57
  Such a waste of power. It's a perfect place to apply
• 8:59
  the gravity principle. Why incur the operational
• 9:02
  expenditure of electricity and the maintenance overhead of
• 9:05
  motors and drive belts when gravity is a constant?
• 9:08
  Gravity is free. It is.
• 9:09
  We pull out those powered belts and install gravity wheel and
• 9:12
  roller conveyors. By calculating the correct
• 9:15
  coefficient of friction and utilizing A slight downward
• 9:18
  declination angle, the cartoons glide effortlessly to the dock.
• 9:22
  It lowers the CapEx, eliminates a failure point and cost nothing
• 9:26
  to run. It's a huge win.
• 9:28
  With the packing area streamlined, we have to tackle
• 9:31
  the massive forklift traffic jams in the main arteries of the
• 9:34
  plant. The trickiest part?
• 9:36
  Human driven vehicles are flexible, but for repetitive
• 9:39
  fixed route transport runs they introduce variability and safety
• 9:42
  risks. Here we introduce the automation
• 9:45
  principle. We replace the manual lift
• 9:47
  trucks in those high traffic predictable loops with an
• 9:50
  automated guided vehicle system or Agvs.
• 9:54
  Now, Agvs sound great on paper. You see them silently gliding
• 9:57
  around modern facilities, but in a chaotic plant where someone
• 10:01
  might drop a pallet splinter or leave a tool cart in the aisle,
• 10:04
  don't those systems fail? They used to.
• 10:06
  Yeah, because I've seen older Agvs that rely on inductive
• 10:09
  loops embedded in the concrete just stop dead in their tracks
• 10:13
  if the floor gets damaged or blocked.
• 10:14
  That was certainly true of older generations that required fixed
• 10:17
  physical guide paths, but modern Agvs and autonomous mobile
• 10:20
  robots or AM Rs use advanced SLAM navigation.
• 10:24
  Simultaneous localization and mapping.
• 10:26
  Right. They utilize onboard Lidar and
• 10:28
  vision systems to map their environment in real time.
• 10:32
  If an AMR encounters a tool cart left in the aisle, it doesn't
• 10:35
  just stop and trigger an alarm it.
• 10:36
  Just routes around it. It calculates A dynamic path
• 10:39
  around the obstacle, ensuring continuous throughput while
• 10:42
  adhering to strict safety protocols.
• 10:45
  They never get fatigued, they maintain precise pacing, and
• 10:48
  they virtually eliminate the traffic bottlenecks.
• 10:51
  OK. We've optimized the horizontal
• 10:53
  movement, but we still have a massive cubic storage problem.
• 10:56
  We've got pallets taking up valuable manufacturing floor
• 10:59
  space. Time for the space utilization
• 11:01
  principle. Yes, the floor is for value add
• 11:04
  manufacturing. The vertical air space is for
• 11:06
  storage. To reclaim that square footage,
• 11:09
  we install an Automated Storage and Retrieval System, or ASRS.
• 11:13
  And ASRS completely transforms the physical footprint of a
• 11:17
  facility. We are talking about engineered
• 11:19
  steel rack structures that can easily reach 80 to 100 feet high
• 11:23
  with incredibly narrow aisles. It's an imposing sight.
• 11:26
  It is down the center of each aisle a computer controlled
• 11:30
  storage and retrieval machine operates.
• 11:32
  It rides on a floor rail and utilizes a top guide channel to
• 11:36
  manage the extreme mass deflection that occurs at those
• 11:39
  heights. The kinematics of those machines
• 11:41
  are wild. They really are.
• 11:43
  The machine accelerates down the aisle while simultaneously
• 11:46
  hoisting the shuttle mechanism vertically.
• 11:48
  It calculates the optimal kinematic curve to reach a
• 11:51
  specific bin location on the 50th tier.
• 11:53
  Smoothly decelerating to prevent load shifting.
• 11:56
  Exactly and automatically depositing or retrieving a multi
• 12:00
  ton pallet with millimeter precision.
• 12:02
  It maximizes storage density and provides real time inventory
• 12:06
  tracking. But wait, an ASRS is great for
• 12:08
  massive pallet loads? What of this factory also
• 12:11
  handles thousands of unique high mix, low volume components?
• 12:16
  Things like specialized fasteners, custom electronic
• 12:18
  relays, or raw plastic pellets for an injection molding wing.
• 12:22
  A very common scenario. You aren't putting loose plastic
• 12:25
  pellets on a palette and you don't need 100 foot crane to
• 12:27
  grab boxes screws. That's a critical distinction.
• 12:30
  Let's tackle the small discrete parts 1st.
• 12:33
  For high density small item storage, we scale down the
• 12:36
  technology and install a mini load ASRS or a horizontal
• 12:40
  carousel. Oh, I love the carousels.
• 12:43
  A horizontal carousel is essentially A continuous
• 12:45
  motorized Oval track of linked bins.
• 12:48
  Instead of a worker wasting time walking up and down static
• 12:51
  shelving aisles searching for a tiny part, which is a huge
• 12:55
  violation of our place in motion principles, the system operates
• 12:58
  on a goods to person paradigm. The computer spins the carousel
• 13:02
  and presents the exact bin directly to the operator's
• 13:05
  picking station. It's incredibly efficient, but
• 13:07
  before management signs off on millions of dollars in capital
• 13:10
  expenditure for 100 foot steel racks, laser guided robots and
• 13:13
  carousels, there is one non negotiable final step.
• 13:17
  You cannot just bolt this equipment to the floor and hope
• 13:19
  your from toe charts were perfectly accurate.
• 13:22
  Absolutely not. The crucial final step is
• 13:24
  computer simulation. Building a digital twin.
• 13:27
  Yes, as an engineer you must build a dynamic computerized
• 13:31
  model of the entire proposed layout.
• 13:34
  You feed the simulation your actual production schedules,
• 13:37
  your machine cycle times, and the kinematic profiles of your
• 13:40
  new Agvs and Asrs. Because if you don't, you might
• 13:43
  just be shifting the bottleneck from one department to another.
• 13:46
  Right precisely. The simulation allows you to run
• 13:49
  weeks of production in a matter of minutes.
• 13:51
  You can identify hidden system bottlenecks, test edge cases,
• 13:55
  and discover timing issues in the virtual world.
• 13:57
  You catch the glitches before they're expensive hardware
• 14:00
  problem. Exactly.
• 14:01
  For instance, the simulation might reveal that during a
• 14:03
  specific shift change, your entire fleet of Agvs will route
• 14:07
  themselves to the charging station simultaneously, totally
• 14:10
  crashing your throughput. You find and fix these logic
• 14:13
  errors in the code long before the millwrights start bolting
• 14:17
  metal to the floor. So by applying these principles,
• 14:20
  unitizing the loads, leveraging gravity, implementing advanced
• 14:23
  automation, utilizing vertical space, and validating it all
• 14:26
  with a digital twin, we've taken a chaotic, failing factory and
• 14:31
  engineered a highly optimized, predictable system.
• 14:34
  We've fundamentally changed the operating physics of the plant.
• 14:37
  No, this walkthrough actually sets us up perfectly for our
• 14:40
  debrief. We like to take the most
• 14:42
  frequently searched online questions about material
• 14:45
  handling and ground the answer straight in the source text.
• 14:48
  Sounds good, let's do it. FAQ number one.
• 14:51
  Earlier I mentioned raw plastic pellets for the injection
• 14:54
  molding wing. The engineering solutions we
• 14:57
  just applied were all focused on discrete countable items.
• 15:01
  O the question is what is the difference between bulk and unit
• 15:06
  handling? It's a distinction that often
• 15:08
  trips up systems design. Everything we've discussed so
• 15:11
  far, pallets, cartoons, engine blocks, falls under unit
• 15:15
  handling. You are moving discrete
• 15:17
  entities. Bulk handling deals with solids
• 15:19
  that are flowable. Think of fine powders like wheat
• 15:22
  flour, pellets like soybeans or lumpy coal.
• 15:25
  You move bulk using completely different gear.
• 15:27
  Right, you can't stack wheat flour on a wooden pallet.
• 15:30
  Instead of an AGV, you might specify a screw conveyor, which
• 15:33
  is essentially A helical shaft in a pipe or bucket elevators
• 15:37
  for vertical lifts. Or pneumatic conveyors that
• 15:39
  suspend powders in a stream of air.
• 15:41
  That leads right into FAQ #2 Why do bulk materials get stuck
• 15:45
  inside storage silos? We've all seen massive
• 15:49
  industrial silos, but extracting material from them reliably is
• 15:53
  notoriously difficult. It comes down to the design of
• 15:56
  the hopper, the sloped bottom of the bin, and the materials flow
• 15:59
  ability. Ideally you want mass flow where
• 16:02
  all the material in the bin moves when the outlet opens.
• 16:05
  But if you miscalculate, you don't get mass flow, you get
• 16:09
  funnel flow. Exactly.
• 16:10
  In funnel flow, only a small vertical channel of material
• 16:14
  directly above the outlet moves, leaving dead zones clinging to
• 16:17
  the walls. To fix it, engineers use flow
• 16:20
  assisting devices. Like vibrating hoppers?
• 16:22
  Vibrating hoppers or Whirlpool devices to break up those
• 16:25
  cohesive bonds. All right, final question FAQ 3.
• 16:28
  When should a factory use a conveyor versus an industrial
• 16:31
  truck? Conveyors are perfect when loads
• 16:33
  are uniform, routes do not vary, movement is continuous point to
• 16:37
  point, and the path is fixed. They're relentless, but they
• 16:40
  lack flexibility. Exactly.
• 16:42
  Industrial trucks like forklifts or Agvs are the go to when you
• 16:46
  need flexible handling along variable or random flow path.
• 16:50
  It's incredible how much rigorous engineering goes into
• 16:53
  simply getting apart from point A to point B.
• 16:56
  Before we wrap up, I want to leave you with a final thought
• 16:59
  to ponder. Consider the invisible
• 17:01
  choreography that surrounds you right now.
• 17:03
  It's everywhere. Almost every physical object you
• 17:06
  own was seamlessly passed from a gravity conveyor to a stretch
• 17:11
  wrapper, to a laser guided robotic vehicle and up into 100
• 17:16
  foot automated storage rack, all without a single human hand
• 17:20
  touching it. It's wild to think about.
• 17:22
  Now, as artificial intelligence merges with these automated
• 17:25
  guided vehicles and automated retrieval systems, how long will
• 17:29
  it be before factories are no longer places where humans use
• 17:31
  machines, but rather giant self-contained robots that
• 17:35
  simply operate themselves? If you haven't already, give us
• 17:38
  a like and follow and please leave us a comment if you enjoy
• 17:40
  the podcast. We love to hear from the folks
• 17:42
  who make this possible.