Best of LinkedIn: Connected Tools & Equipment CW 43/ 44

00:00:00: Provided by Thomas Allgaier and Freeness, based on the most relevant LinkedIn posts about connected tools in equipment.

00:00:05: In calendar weeks forty-three and forty four.

00:00:08: Freenes is a B to B market research company that supports enterprises in the power tool sector with building feature by feature competitive intelligence.

00:00:16: That shows exactly how their product stacks up against the competition.

00:00:19: you can find more info in the description.

00:00:21: So today we're setting out to explore some of the top trends across linkedin regarding Connected Tools & Equipment

00:00:27: right?

00:00:27: Yeah Exactly.

00:00:29: And we are really looking at this absolute convergence of maintenance protocols and physical tooling moving from foundational strategies to, you know high-take vibration analysis the scaling of industrial AI.

00:00:41: Then then the actual connected physical tools reshaping job sites.

00:00:46: Okay Let's unpack this because before we can even begin to talk about advanced artificial intelligence in wireless sensor networks We really have to establish a hard truth

00:00:55: which is

00:00:55: That's slapping new sensors on a broken maintenance system just doesn't work.

00:00:59: Well,

00:01:00: totally is it complete waste of capital?

00:01:01: Right and to put some real financial weight behind this I've been get off recently highlighted Some really fascinating numbers from the US Department Of Energy.

00:01:10: Okay well they fine.

00:01:11: They found that preventive maintenance program.

00:01:14: You know where you're routinely greasing bearings And swapping filters on schedule.

00:01:19: That saves A facility roughly twelve To eighteen percent over simply running equipment until It blows up

00:01:24: Which makes sense yeah.

00:01:25: But here is the kicker.

00:01:27: A predictive maintenance program, which has driven by advanced data adds another eight to twelve percent in savings but only when it's measured against a preventive program that actually functioning properly.

00:01:39: That sequencing rule means everything and the mechanics behind explain exactly why.

00:01:44: Right Like if you install highly sensitive predictive sensor system on a plant that still culturally runs its machines until breaking point all you've really done for just an expensive alarm clock.

00:01:55: Yeah, an alarm clock that everyone ignores.

00:01:57: Exactly!

00:01:58: The system detects a micro fracture in the gear and an alert lands into manager's inbox.

00:02:04: but if you don't have dedicated maintenance planner or lack of spare parts it is useless.

00:02:10: Oh...or If You Can't Get Production Team to Agree To A Scheduled Window To Actually Open The Machine Up Yes

00:02:16: That Predictive Data Is Entirely Useless.

00:02:19: Julio Panzerra made a brilliant observation about this reality.

00:02:22: What do you say?

00:02:23: Well, after thirty years in the industry working across these sprawling steel rolling mills and food processing plants He says he still sees companies pouring capital into AI software.

00:02:32: Right that shiny new toy

00:02:33: right.

00:02:34: yet at the end of the day It's the technician walking the floor who actually catches the failure just by listening to the pitch or the bearing noise.

00:02:42: The problem isn't the technology it's the method.

00:02:45: vibration data without context is just noise.

00:02:49: But I have to push back a little here.

00:02:51: Okay, go for it.

00:02:52: It isn't the temptation for management To just buy The shiny new sensor and hope it acts like A magic band-aid on a broken leg.

00:02:58: Oh absolutely this!

00:02:59: It Just feels so much easier to Buy a tech solution than to Fundamentally rewrite how a plant operates.

00:03:05: And This ties perfectly into an observation from Stuart J Lawrence.

00:03:09: What was his take?

00:03:10: He pointed out a very common trap Just because a machine is running Does not mean it Is healthy.

00:03:16: Ah yes the illusion of health.

00:03:19: Exactly, mechanically speaking you can have an induction motor spinning perfectly at eighteen hundred RPM while bearing inside is slowly collapsing.

00:03:28: The machine running but it's on a countdown to unplanable failure.

00:03:32: That's ticking time bomb.

00:03:33: And that brings us into ultimate human element in this foundation which Kishore Prabhakar Kaplay articulated perfectly.

00:03:40: What was the human element there?

00:03:42: He noted that true test maintenance excellence isn't doing more maintenance It's knowing when to stop production.

00:03:50: Wow, yeah.

00:03:52: Trusting your team enough to take a short-term hit... ...to avoid massive business disaster?

00:03:57: Exactly!

00:03:57: Yeah

00:03:58: it is huge leadership challenge

00:03:59: which has remarkably heavy decision make on the plant floor.

00:04:02: but that human element of knowing When To Stop Production connects directly to diagnostic data.

00:04:08: right The Data That Gives Teams Confidence To Push That Stop Button.

00:04:11: Yeah and physical evidence can be incredibly counterintuitive if you don't understand the underlying physics.

00:04:18: Give me an example.

00:04:19: So Victor Appalon shared this crazy case, picture a towering furnace exhaust fan that suddenly started exhibiting severe one-x vibration.

00:04:28: Okay so one x That means the vibration is occurring at the exact running speed of the machine right?

00:04:33: Spot on!

00:04:35: The fan spins at fifteen hundred RPM.

00:04:37: it's vibrating exactly fifteen hundred times per minute.

00:04:40: Usually

00:04:41: there's heavy spot throwing off center every single rotation.

00:04:45: So centrifugal force is just ripping the assembly around?

00:04:48: Exactly.

00:04:49: The initial engineering recommendation was to weld heavy balancing weights onto the fan wheel, to counteract it.

00:04:55: Seems logical

00:04:56: Right But when the reliability team looked at maintenance history they realized this severe unbalance started right after a preventive cleaning task.

00:05:04: Wait so the maintenance itself caused that issue.

00:05:07: Yes!

00:05:08: The Cleaning was incomplete.

00:05:10: Some of the massive steel blades still had baked on deposits, causing a radically uneven mass distribution.

00:05:16: Oh

00:05:16: wow!

00:05:17: The root cause wasn't an unbalanced structure it was just a terrible cleaning job.

00:05:22: they powerwashed it properly and the vibration dropped to a smooth baseline no welding required.

00:05:27: Here's where it gets really interesting.

00:05:29: Because the data doesn't always violently trip a breaker like

00:05:32: that, right?

00:05:32: No sometimes its really subtle

00:05:34: Right!

00:05:35: QUN Sun shared an incredible example of a subtle vibration trend.

00:05:40: Picture a heavy rotating blade system Where the spectral amplitude you know The intensity is sitting at very comfortable

00:05:46: point three Okay totally normal Yeah.

00:05:48: But then over weeks It slowly creeps up to point six Then one point seven and finally hits two point four.

00:05:55: So a slow climb.

00:05:56: Right, and none of those early numbers were high enough to trigger a flashing right alarm in the software.

00:06:01: but the signal wasn't the threshold itself.

00:06:03: it was the trend.

00:06:05: that's low steady-climb exposed a microscopic fatigue crack propagating inside the rotating blade.

00:06:11: if they just waited for the software across a generic danger threshold That blade would have eventually sheared off at full speed

00:06:18: which is terrifying Yeah And spotting a tread like that requires such sophisticated analysis.

00:06:24: Sometimes you have to filter out the noise of entire factory.

00:06:26: define the whisper.

00:06:27: Right.

00:06:28: Mustafa Durson documented a fascinating case using envelope analysis on motor drive end bearing, sixty three fourteen bearing.

00:06:36: Okay.

00:06:37: Now overall vibration was fine.

00:06:40: it stayed safely below one millimeter per second RMS.

00:06:43: Wait, I need to pause you there.

00:06:44: RMS.

00:06:45: Root means square?

00:06:46: If it's below one millimeter per second doesn't that mathematically mean the machine is perfectly

00:06:51: healthy?"?

00:06:51: You'd think so!

00:06:52: A novice analyst would walk right past it.

00:06:54: but envelope analysis acts like an extreme audio filter.

00:06:58: It strips out the heavy low-frequency hum of the motor To isolate sharp high-pitched impacts on a tiny defect

00:07:06: Like finding needle in haystack

00:07:08: Exactly.

00:07:09: And by doing this, Derson's team caught a tiny nineteen point two Hertz peak that matched the defect frequency of the bearings cage.

00:07:17: Nineteen point two hertz meaning that specific piece of bearing was failing almost twenty times every single second completely hidden beneath normal vibration.

00:07:26: and what is wild isn't just for rotating equipment anymore.

00:07:30: Will Salmon applied vibration analysis to complete static assets.

00:07:35: He was checking a power transformer for internal looseness, or else wetty electric.

00:07:40: Wait how do you measure vibration on a machine with no moving parts?

00:07:44: It comes down to electromagnetism.

00:07:46: the alternating current moving through a transformer creates these powerful oscillating magnetic fields.

00:07:52: oh I see

00:07:53: yeah it physically pulls and squeezes the heavy steel core one hundred twenty times per second.

00:07:57: if internal bolts loosen up those steel plates start buzzing.

00:08:01: Salman proved this internal mechanical looseness actually changes the external vibration patterns.

00:08:07: That is wild!

00:08:08: Even a solid block of steel is telling a mechanical story... Exactly And you know, listening is quite literally the right word because condition monitoring is expanding into the visual spectrum with acoustic imaging cameras.

00:08:22: Oh really?

00:08:22: Yeah

00:08:23: Said Ali Azimi demonstrated it.

00:08:26: He aimed an acoustic camera at running industrial fan and belt drive assembly.

00:08:30: Okay.

00:08:31: These cameras use highly sensitive microphones to map ultrasonic friction and visually overlay it onto a screen without shutting the equipment down.

00:08:40: they pinpointed A sixty one point five decibel sound source right at The Bell Guard

00:08:45: so, They can physically see the Sound of the Friction

00:08:48: exactly

00:08:49: now you still need traditional vibration analysis To confirm the root cause but It instantly narrows Down Exactly where?

00:08:56: The engineers Need to focus

00:08:58: But if we step back for a second This raises a crippling logistical issue.

00:09:03: if we're capturing nineteen point two Hertz anomalies and sixty one point five decibel sound sources across thousands of machines How does the human team process all that without just?

00:09:14: Drowning in data.

00:09:16: Well, they can't

00:09:16: made completely drown.

00:09:18: And that data bottleneck is exactly what's driving the massive rise Of industrial AI.

00:09:24: The volume of this data requires AI to sort the signal from the noise

00:09:28: right And to give you an idea of the sheer scale of capital pouring into this, David Juergens recently discussed how Tractian just raised a hundred and thirty million dollar series D at one point four billion valuation.

00:09:41: That is massive.

00:09:42: It's

00:09:42: huge, and that capital is focused on expanding their physical AI partner network deploying predictive hardware across thousands of facilities.

00:09:50: Right but thats a very Hardware intensive approach.

00:09:52: There s completely different software only approach gaining traction too Oh yeah?

00:09:56: Yeah Marvin C Blackman shared a Predictive Maintenance Solution Built On Iconics Genesis V-Eleven.

00:10:02: Instead of bolting thousands Of new sensors onto motors it scores the health of every asset using data that already exists inside The plants.

00:10:09: variable frequency drives

00:10:11: Oh wait, a VFD just regulates the power going to the motor right?

00:10:14: How does that act as a sensor.

00:10:16: Well...to regulate the speed of EFT has constantly measure electrical resistance voltage and current draw.

00:10:22: So it acts.

00:10:23: is virtual torque beater Blackman software pulls those run hours faults in electrical loads directly from the VFDS over existing network Zero new physical sensors added.

00:10:33: That's incredibly efficient.

00:10:34: By the way if you are finding this breakdown useful make sure subscribe so don't miss our future deep dives Definitely.

00:10:40: But you know, whether a plant uses a sprawling new sensor network or taps into existing VFD data.

00:10:47: Andy Sharples made a crucial point.

00:10:48: detection is actually the easy part.

00:10:50: The hard part is minimizing false positives.

00:10:53: Oh

00:10:53: for sure.

00:10:54: Sharples argues For a glass box AI approach so engineers can Actually see the reasoning behind an alert.

00:11:00: let me ask You this if An AI cries wolf just two Or three times on A plant floor does it instantly ruin?

00:11:05: The credibility of the entire multi-million dollar program?

00:11:08: oh It absolutely destroys it.

00:11:10: The engineers will just put black tape over the dashboard monitor.

00:11:13: Right.

00:11:14: And

00:11:14: preventing those false positives often comes down to data quality.

00:11:18: Colton Bach shared how machine saver captures full time wave form vibration data, over Lora WAN instead of compressed summaries.

00:11:26: and Laura WAN is crucial because standard factory Wi-Fi just bounces off steel tanks right?

00:11:31: Exactly it punches through industrial infrastructure giving me AI that complete uncompressed waveform, every single peak in Valley drastically reduces the false alarms.

00:11:42: But getting that high quality network approved by corporate IT is a whole different nightmare.

00:11:46: John Webb shared a really painful cautionary tale from The Sales Side.

00:11:50: Oh no what happened?

00:11:51: He had A four hundred twenty five thousand dollar wireless deal verbally committed.

00:11:56: The plant was bleeding, three point two million dollars a year in unplanned downtime.

00:12:01: But IT said they needed sixty to ninety days to evaluate the network protocols.

00:12:05: The dreaded IT black hole?

00:12:06: Yep!

00:12:07: The deal completely died during the wait and the plant paid what Web calls a Wait-and-See tax roughly two hundred sixty seven thousand dollars per month an unplanned down time while waiting for security review.

00:12:18: That is devastating.

00:12:20: but let's assume you win that battle.

00:12:22: get the IT approval in Generative Verified Work Order.

00:12:25: that predictive data ultimately has to result in a physical action on the job site, right?

00:12:31: Yes.

00:12:32: The digital intelligence has to be matched by the physical capability and ergonomics of the tweets

00:12:36: themselves.".

00:12:37: And we're seeing an incredible push there!

00:12:40: Momin Nawasreh & Makmood Lofty highlighted Hilti UAE's power-up.

00:12:44: twenty-twenty six event.

00:12:46: they launched seven brand new Neuron heavy duty cordless tools

00:12:52: Right, breaking away from the heavy cords.

00:12:54: Adam Purcell actually tested a new Neuron AG- Thirty dash twenty two grinder specifically on a railway track welds

00:13:00: Which is brutal environment.

00:13:02: Exactly no fuel or trailing leads required.

00:13:05: and Ben B tested the new SR ten reciprocating saw noting it cuts faster with visibly lower vibration than older SR six model.

00:13:13: And that vibration reduction huge for worker safety.

00:13:16: It's not just cutting either.

00:13:18: Frank Fernandez showcased the max rev light and turbo driver, running a hundred fastener capacities on single power like compressor platform.

00:13:25: So what does this all mean?

00:13:27: When we tie these ergonomic tools back to digital maintenance data.

00:13:30: Well it means speed of physical repair is bottlenecked.

00:13:33: by searching for equipment

00:13:35: Geel Kerscher posted about the Walter Tool Station C-Line TX-七.

00:13:39: fifty.

00:13:40: It's an automated locker system that gives access to nine hundred eighty seven controlled compartments in under ten seconds To cut search time,

00:13:47: but we still have to remember the human at the end of the tool.

00:13:49: very true Stefan Taylor brought up a brilliant point on bolting assurance.

00:13:54: You can calibrate a torque tool perfectly.

00:13:56: proving The equipment is fine But it doesn't prove the operators competence.

00:14:00: you still need certified training like ECITB MGI to close the gap.

00:14:05: It's like having a perfectly tuned AI-monitored Formula One car.

00:14:09: Doesn't matter if the driver doesn't know how to take the corners, The human expertise has to match the tech!

00:14:14: Exactly

00:14:15: Well If you enjoyed this episode new episodes drop every two weeks.

00:14:18: Also check out our other editions on robotics smart manufacturing and industrial AI And digital construction.

00:14:24: Thank You so much for joining us.

00:14:25: Don't forget to subscribe

00:14:27: As we wrap up.

00:14:28: think about as tools become increasingly digitized how long until the tool itself can automatically detect a novice user's improper technique and lock it self down before they make critical error on multi-million dollar asset.