Business of Automation

IRSC experts detail ANSI R15.06/ISO 10218 updates, ISO 21423 interoperability, and safety challenges for dynamically stable robots.

29 min Listen to the original
Robot Safety StandardsANSI R15.06ISO 10218ISO 21423ISO 25785Dynamically Stable RobotsAI Functional SafetyIRSC

Summary

Aaron Prater and Carol Franklin discuss the International Robot Safety Conference (IRSC) and recent updates to industrial robot safety standards. They explain that the 2025 editions of ANSI R15.06 and ISO 10218 are technically identical, with R15.06 Part 3 imposing specific change management and training responsibilities on end-users. The speakers introduce ISO 21423, a new standard for mobile robot interoperability developed from a Mass Robotics specification, and ISO 25785, which addresses safety for dynamically stable robots like humanoids. They note that AI is becoming critical for managing stability in these advanced robots, leading to new joint ISO/IEC functional safety standards.

Episode sections

  1. 0:00 – 2:57

    Introduction and IRSC Context

    Aaron Prater and Carol Franklin discuss the International Robot Safety Conference (IRSC) in Detroit, highlighting its role in educating attendees on robot safety standards. They note that A3 has been involved in industrial robot safety since the mid-1980s, spanning 40 years.

  2. 2:57 – 9:36

    R15.06 vs R15.08 and ISO 10218 Evolution

    The speakers explain that R15.06 is the ANSI standard for industrial robot safety, originally developed in 1986, while R15.08 covers industrial mobile robots. They detail how R15.06 was adopted into ISO 10218, with the 2025 editions of R15.06 and ISO 10218 being technically identical through national adoption.

  3. 9:36 – 12:37

    Industrial vs. Service Robot Environments

    Carol Franklin distinguishes between industrial robots, which operate in controlled environments with trained adults, and service robots, which may face public access from unvetted or untrained individuals, including children, making safety standards more complex.

  4. 12:37 – 16:56

    R15.06 Part 3: User Responsibilities and Change Management

    The discussion covers R15.06 Part 3, which addresses user responsibilities such as worker training and change management. Unlike ISO standards that only regulate manufacturers, US ANSI standards include requirements for end-users to manage system modifications and risk assessments.

  5. 16:56 – 21:31

    ISO 21423: Interoperability for Mobile Robots

    A new ISO standard, ISO 21423, is introduced to address interoperability issues among mobile robots from different vendors. It originated from a Mass Robotics specification and aims to prevent productivity-killing standoffs between machines by establishing international communication protocols.

  6. 21:31 – 25:29

    ISO 25785: Dynamically Stable Industrial Mobile Robots (DSIMR)

    The speakers discuss ISO 25785, a standard in development for dynamically stable industrial mobile robots (DSIMR), including humanoids and quadrupeds. These robots require active actuator management for stability, meaning an emergency stop can create new hazards by causing the robot to fall.

  7. 25:29 – 29:00

    AI and Functional Safety Standards

    The conversation touches on the role of AI in managing DSIMR stability and the development of a new joint ISO/IEC standard on AI and functional safety. This emerging standard will likely be referenced by the DSIMR standard to address how AI impacts safety systems.

Claims worth verifying

A3 has been involved in industrial robot safety since the mid-1980s, specifically starting with R15.06 development in 1982 and publishing it in 1986.

Carol Franklin ·2:11

How to verify: Check A3 (formerly RIA) historical records or the publication date of ANSI/RIA R15.06-1986.

The 2025 edition of R15.06 is technically identical to the 2025 edition of ISO 10218, with no national deviations.

Carol Franklin ·6:17

How to verify: Compare the technical requirements sections of ANSI/RIA R15.06-2025 and ISO 10218-1:2025/ISO 10218-2:2025.

ISO 10218 Part 1 covers the industrial robot arm itself, while Part 2 covers the integrated system or cell.

Carol Franklin ·4:50

How to verify: Review the scope sections of ISO 10218-1 and ISO 10218-2.

ISO documents are not permitted to place requirements on the user of a machine, only on the manufacturer.

Carol Franklin ·13:04

How to verify: Check ISO/IEC Directives, Part 1, which governs the development of ISO standards.

The update cycle for ISO 10218 took 13.5 years, with COVID adding one year and the Haas consultant situation adding 1.5 to 2 years.

Carol Franklin ·11:54

How to verify: Review ISO project timelines or A3 conference records for the ISO 10218 revision history.

ISO 21423 is the new standard code for interoperability of industrial mobile robots, derived from a Mass Robotics specification.

Carol Franklin ·19:49

How to verify: Search for ISO 21423 publication or status on the ISO website.

ISO 25785 is the standard number for dynamically stable industrial mobile robots (DSIMR), including humanoids and quadrupeds.

Carol Franklin ·22:05

How to verify: Search for ISO 25785 project status on the ISO website.

For dynamically stable robots, pushing the emergency stop removes power to actuators, which can cause the robot to fall and create a new hazard.

Carol Franklin ·23:47

How to verify: Review safety engineering principles for dynamically stable robots or draft ISO 25785 content.

ISO and IEC are jointly developing a standard on AI and functional safety through Joint Technical Committee 1 (JTC1).

Carol Franklin ·26:36

How to verify: Check ISO/IEC JTC1 project lists for standards related to AI and functional safety.

R15.06 Part 3 and Canada's Z434.1 are harmonized with the same technical requirements for user safety.

Carol Franklin ·14:11

How to verify: Compare the text of ANSI/RIA R15.06-2025 Part 3 and CSA Z434.1.

Do your own research

  • Verify the text of ANSI R15.06 Part 3 (2025 edition) to confirm specific end-user change management requirements.
  • Review ISO 21423 draft or final text to understand interoperability protocols for mobile robots.
  • Check ISO 25785 status and scope for dynamically stable industrial mobile robots (DSIMR).
  • Examine joint ISO/IEC standards on AI and functional safety for references to DSIMR applications.
  • Assess Mass Robotics publications regarding the origin and technical details of ISO 21423.
  • Consult IRSC archives or proceedings for detailed presentations on R15.06 vs R15.08 distinctions.
  • Review manufacturer compliance documentation for ISO 10218 alignment with ANSI R15.06.
Full transcriptexpand
0:00
Unidentified speaker: What happens when you push the E stop has to be different.
0:03
Unidentified speaker: Remove all power to the actuators, and you're actually creating a new hazard.
0:06
Unidentified speaker: Carol, Director of Standards Development here at A3, how's it going today?
0:10
Unidentified speaker: In the mobile robot, it takes its hazard zone with it.
0:13
Unidentified speaker: You end up having to treat the entire facilities with the same amount of care that you put into your industrial robot cells.
0:20
Unidentified speaker: Two machines from different vendors would block each other and then they would engage in the standoff.
0:25
Unidentified speaker: A human has to go figure it out, and that's just not a sustainable way to be.
0:29
Unidentified speaker: In the industrial environment, which you've got are adults able to be trained.
0:33
Unidentified speaker: In the surface robot side, you might have completely public access to this machine, totally unvetted, totally untrained.
0:40
Unidentified speaker: When you start to move outside of the industrial space, that's actually a harder problem.
0:44
Unidentified speaker: It's not actually possible for a human to make all of the tiny adjustments to keep the thing stable.
0:49
Unidentified speaker: Part one is the manufacturer of the robot, part two is the integrator, part three is the user.
0:56
Unidentified speaker: So does that new piece of equipment change how the whole system behaves?
1:00
Unidentified speaker: And if so, does that change the risk profile that your workers are supposed to?
1:11
Unidentified speaker: Welcome to the business of automation, live from Detroit in the IRSC, the International Robot Safety Conference.
1:17
Unidentified speaker: I'm Aaron Prater, Director of Market Intelligence here at A3, the Association for Advancing Automation.
1:23
Unidentified speaker: And IRST is a very special conference to me because it was the very first conference I attended at A3.
1:31
Unidentified speaker: So it has a special place in my heart.
1:34
Unidentified speaker: And the person I met there that really actually helped me actually get my A3 foot underneath me was Carol Franklin, who is joining us today.
1:43
Unidentified speaker: Carol, Director of Standards Development here at A3.
1:46
Unidentified speaker: How's it going today?
1:47
Unidentified speaker: Going great, Aaron.
1:48
Unidentified speaker: And thanks for having me on the show.
1:50
Unidentified speaker: Great to have you.
1:52
Unidentified speaker: So IRSC is this week.
1:54
Unidentified speaker: What is going to be the big takeaways that you're looking for for all the uh attendees that are coming to Detroit this week?
2:01
Unidentified speaker: So IRSC is uh really important uh from the standpoint of educating people about what it really takes to keep people safe around robots.
2:11
Unidentified speaker: A3 has been involved in industrial robot safety since the mid-1980s.
2:17
Unidentified speaker: So that's 40 years now.
2:19
Unidentified speaker: And so uh we started holding IRSC to help unpack what is in the industrial robot safety standard.
2:28
Unidentified speaker: And so uh we're still doing that, but also since then we've had mobile robot safety standards and other standards as um robotics has expanded and moved into new environments, so does the need for standards.
2:41
Unidentified speaker: So we have a lot, we have a lot to share this week.
2:44
Unidentified speaker: I remember when I went to my very first IRSC, the only standard we really had was 15.06 on ARM.
2:51
Unidentified speaker: And there was talk about doing mobile robots at the time, and that kicked off the whole R15.08.
2:57
Unidentified speaker: I was I was so happy to actually join that that effort.
3:01
Unidentified speaker: So we're throwing these numbers around R15.06, R15.08.
3:06
Unidentified speaker: We're probably gonna throw some ISO numbers into this.
3:09
Unidentified speaker: So what what is the difference between just to start out, 15.06 versus 15.08?
3:16
Unidentified speaker: Yeah, so those numbers um that start with R15 are our ANSI standards.
3:22
Unidentified speaker: So that's American National Standards Institute.
3:24
Unidentified speaker: So those are American national standards, and so those were developed specifically in the US for the US market.
3:32
Unidentified speaker: R1506 is the great granddaddy of them all.
3:36
Unidentified speaker: So that's the the very first voluntary industry consensus standard on robot safety that was that was developed 40 years ago, starting in um 1982 and published in 1986.
3:49
Unidentified speaker: And that was by uh the RIA, which is uh the former name of A3.
3:54
Unidentified speaker: So um an R1506 is for uh safety requirements for industrial robots and robot systems.
4:02
Unidentified speaker: Uh so we first started it as an ANSI standard, and then a few iterations in we realized that it needed to become an ISO standard because an ANSI standard really only holds sway within North America.
4:18
Unidentified speaker: And ISO, though, you know, a lot of our companies are multinational uh companies in this industry, and so they really saw a need for international consistency across borders.
4:30
Unidentified speaker: So we took the 99 edition of R1506, which was a very popular, very well thought-of safety standard at the time, and we offered that to ISO as a seed document for an ISO standard on the same topic.
4:44
Unidentified speaker: And after a number of years of development, that the result was the 10 to 18.
4:50
Unidentified speaker: And so 10 to 18 breaks up the chunks of information into part one, which is for the industrial robot itself, which is the arm, robot arm, but the arm itself, having no end effector, can't do any work.
5:04
Unidentified speaker: So it has to be integrated into a system or industrial robot system or application.
5:09
Unidentified speaker: And that is the subject of part two.
5:12
Unidentified speaker: So part one is for the robot arm itself, part two is for the integrated system, uh whole cell, which includes the application, includes the workpiece.
5:21
Unidentified speaker: And so um part one and part two is in the ISO system, ISO 10 to 18.
5:27
Unidentified speaker: Then because we had um been very, you know, we offered R1506 as the starting point for 10 to 18, and then also the US was very strongly involved in developing 10 to 18.
5:39
Unidentified speaker: So we felt that would serve the needs of our market, and so we did what's called a national adoption.
5:44
Unidentified speaker: So the 2012 edition of R1506 was a national adoption of 10 to 18 of R 2011, and so that's really been that's been our approach since the 2011-2012 edition.
5:58
Unidentified speaker: So uh now we have recently the ISO group has published a new 10 to 18 parts one and two, and that was published in 2025, and we managed to do the national adoption in the same calendar year.
6:12
Unidentified speaker: So the R1506, the newest edition of R1506, is also 2025.
6:17
Unidentified speaker: So the R1506 part one and two is functionally uh is technically identical.
6:24
Unidentified speaker: We did what's called a uh identical national adoption, so it's not we had no national deviations, which means we made no changes to the technical requirements.
6:34
Unidentified speaker: So, if in other words, if you are following R-1506 of 2025, that means you're also consistent with the requirements of 10 to 18 of 2025.
6:44
Unidentified speaker: And so, you know, in the US, if you're following the R-1506, that means you're following the same rules as in Canada, the Z434, which is also a national adoption, a Canadian national adoption of 10 to 18, which is the same as whatever it is in Germany and Japan and Korea.
7:01
Unidentified speaker: So all of those countries around the world um do whatever in their local parlance is a national adoption.
7:09
Unidentified speaker: And in so they're all using 10 to 18 based standards.
7:13
Unidentified speaker: And what that means is when you're a company that wants to move and do business in all these different markets, um, you really only have to follow one set of rules for safety, and that's the 10 to 18 based national standards.
7:26
Unidentified speaker: And so that's R1506.
7:28
Unidentified speaker: R1508 adds um mobility into the equation.
7:33
Unidentified speaker: And so R1508 is safety requirements for industrial mobile robots.
7:38
Unidentified speaker: And that one little word, mobile, really adds a tremendous amount of complexity to the requirements because in the R1506 and 10 to 18 framework, um, I think most people are familiar with you've got this robot cell.
7:53
Unidentified speaker: And within the robot cell, you're very carefully controlling everything.
7:58
Unidentified speaker: You're controlling what goes into that cell, what goes out of that cell, and what the robot can or cannot do while that thing is in the cell.
8:06
Unidentified speaker: So if a human goes into the cell and if it's got an interlocked physical guard, the robot system will shut down.
8:13
Unidentified speaker: It will go into a protective stop.
8:15
Unidentified speaker: So um that's to protect the human from hazardous motion.
8:19
Unidentified speaker: On the other hand, if you're passing in uh uh raw material into the cell, that's okay.
8:24
Unidentified speaker: The robot can keep working on it because that's the job it's supposed to be doing.
8:28
Unidentified speaker: So um, so you're very carefully controlling how that cell is is um designed, is built, and uh access to it.
8:38
Unidentified speaker: And so that that's that's in the um, for lack of a better term, fixed-in-place robot arm.
8:46
Unidentified speaker: The mobile robot, though, can move all around the facility where it's being deployed.
8:51
Unidentified speaker: And what happens when the mobile robot moves through the facility is it takes its hazards with it.
8:57
Unidentified speaker: So the hazard zone, which in the robot cell is protected, the person is protected by either physical guards or a system of sensors or what have you.
9:06
Unidentified speaker: In that hazard zone for the industrial robot that has a defined location, uh, that hazard zone is very carefully controlled.
9:14
Unidentified speaker: With a mobile robot, it takes its hazard zone with it.
9:17
Unidentified speaker: And so you end up having to treat the entire facility that the mobile robot can access with the same amount of care that you put into your industrial robot cell.
9:28
Unidentified speaker: And so um, those are the two biggies, the robot arm and the mobile robots.
9:33
Unidentified speaker: And that's in the US or ANSI side.
9:36
Unidentified speaker: Now, in the ISO side, we have, and I should say also it's also those are both industrial.
9:41
Unidentified speaker: So in the ISO side, we make a distinction between industrial and service.
9:46
Unidentified speaker: And the main distinction between those two is the environment in which that machine is deployed.
9:51
Unidentified speaker: And so in the industrial environment, what you've got are adults.
9:54
Unidentified speaker: So you've got legal adults, working capable adults who are generally speaking the size of adults and uh able to be trained in how to safely interact with these machines.
10:05
Unidentified speaker: In the service robot side, instead, you might have um completely public access to this machine, you know, totally unvetted, totally untrained, and in some cases, maybe not fully capable of being trained because, for example, it might be a small child that is not really capable of understanding what they need to understand in order to interact safely with a robot system.
10:26
Unidentified speaker: And so in an industrial setting, you wouldn't find small children there because they're they are prevented from access.
10:32
Unidentified speaker: And so when you start to move outside of the industrial space, that's actually a harder problem because in the industrial space, like I said, it's more controlled access and the people are are trainable and and um capable of doing the work and uh generally a physical size um adult.
10:51
Unidentified speaker: So then when you get into the service robot world, that becomes a much more difficult task.
10:56
Unidentified speaker: And so um it's actually a lot more complicated on the ISO side.
11:00
Unidentified speaker: It's fascinating on how standards are developed.
11:03
Unidentified speaker: As you said, a lot of these robot safety standards started uh here in the United States.
11:09
Unidentified speaker: We went through ANSI, ISO, the international group, sees that the globe needs this.
11:15
Unidentified speaker: So they take this that that seed document, as you talked about, and turned it into an international standard.
11:22
Unidentified speaker: And now there's this symbiotic relationship now between the two standards, even though they're they are now the same.
11:28
Unidentified speaker: It's just how do you place that in?
11:30
Unidentified speaker: And I think that's something that's really fascinating.
11:32
Unidentified speaker: Of when we talk about standards creation, a lot of folks always are like, who should we start with and who should we go through?
11:38
Unidentified speaker: And sometimes it might just be just get started and get something moving because this takes a long time.
11:46
Unidentified speaker: This last update to ISO 1028 took how long?
11:51
Unidentified speaker: Um 13 years.
11:54
Unidentified speaker: 13 and a half years, actually.
11:57
Unidentified speaker: Um, although to be fair, part of that was in uh COVID.
12:01
Unidentified speaker: I would say COVID added a year, and the Haas consultant situation added one and a half to two years.
12:08
Unidentified speaker: But even so, um, yeah, it it was it was quite a long time.
12:12
Unidentified speaker: It was a long time coming.
12:14
Unidentified speaker: But it is just so fascinating just how much work does go into this, as you listed out.
12:19
Unidentified speaker: I mean, so so many things could derail this work, but it's such important work, and I know a lot of the folks put a lot of work into this, so it's amazing to see it.
12:28
Unidentified speaker: Now, one thing that's really interesting though is we are starting to see something called the part threes of both 15.06 and .08.
12:37
Unidentified speaker: And we know ISO does not really take these part threes due to some other reasons.
12:43
Unidentified speaker: Can you explain what these new part threes?
12:45
Unidentified speaker: Because part one, as you mentioned, is the manufacturer of the robot, part two is the integrator, part three is who, Carol?
12:54
Unidentified speaker: The user.
12:55
Unidentified speaker: So the user.
12:57
Unidentified speaker: Yeah, ISO documents are only permitted to place requirements on the manufacturer of a machine and not on the user.
13:04
Unidentified speaker: And so um, in our case, we have multiple levels of manufacturer because we've got the robot manufacturer, and that's who gets part one.
13:11
Unidentified speaker: And as you say, we've got the system integrator.
13:14
Unidentified speaker: So we're looking at the whole system as a completed machine, and the system integrator is the manufacturer of that completed machine or system, and so that's part two.
13:23
Unidentified speaker: But now um ISO documents are not permitted to place requirements on the user.
13:28
Unidentified speaker: So it's actually up to each country then to create our user requirements.
13:34
Unidentified speaker: So in the US, we did develop the part three for R1506, and so we nationally adopted 10 to 18 part one as R1506 part one, 10 to 18 part two as R1506 part two, but then there was no ISO document equivalent for part three because ISO documents don't do that.
13:52
Unidentified speaker: And so we then in the US, we developed the part three entirely sort of homegrown.
13:58
Unidentified speaker: Although we solicited and had very good involvement from our Canadian friends, and they also, I would say that the part threes, our our goal was to have consistency in North America.
14:11
Unidentified speaker: So the part threes in in Arctic06 in the US and Z434 in Canada, again, are harmonized with each other.
14:19
Unidentified speaker: They have the same technical requirements because we worked very closely together.
14:22
Unidentified speaker: Our two committees worked very closely together to develop that part three.
14:26
Unidentified speaker: And so I think um in the US we call it part three.
14:29
Unidentified speaker: In Canada, they call it something else.
14:32
Unidentified speaker: Sorry, Z434.1 or something like that.
14:35
Unidentified speaker: But it is similar to our part three in the sense that it's part of the same Z434 family, and it's the part that we in the North America develop together because ISO documents don't.
14:48
Unidentified speaker: And so the part three is essentially if you, the user of an industrial robot system, um, great, you have an integrator who has created this system for you and they have said they comply with R1506, part one and part two.
15:04
Unidentified speaker: Yay, they hand you the keys and now you own it.
15:07
Unidentified speaker: And so um, how do you, as the user, make sure that your workers did get the training that they need to understand how to approach this machine and what they can and can't do safely around it?
15:20
Unidentified speaker: Um, and also how do you manage the system's change management?
15:25
Unidentified speaker: So if you uh end up having to change the light curtain or something, you do you need a new risk assessment?
15:32
Unidentified speaker: Does that new piece of equipment that you put in to replace something that was broken or whatever, um, does that change how the whole system behaves?
15:40
Unidentified speaker: And if so, does that change the risk profile that your workers are exposed to?
15:45
Unidentified speaker: And so um the part three talks about how do you do change management?
15:50
Unidentified speaker: When do you need to conduct a new risk assessment?
15:53
Unidentified speaker: And um, you know, and when potentially do you need to upgrade or update the rest of the system because some kind of change has caused the system to no longer be as safe as you want it to be.
16:05
Unidentified speaker: I would encourage our end users out there to download, go get this from our A3 website.
16:12
Unidentified speaker: It's a great new ad, and it does it, it really does help guide you through where as an end user you need to take your safety because it, as you said, Carol, the manufacturers have done their job on safety, the integrators now have done their job on safety.
16:28
Unidentified speaker: It's your job to maintain that now going forward.
16:31
Unidentified speaker: So I think it's it's a great new document.
16:34
Unidentified speaker: Um, and I highly encourage if you're an end user, visit automate.org and download that.
16:40
Unidentified speaker: Uh it will definitely help your operations.
16:43
Unidentified speaker: Now, the next topic I wanted to talk to you about with Carol is for the first time, we have done a non-safety standard with ISO on interoperability.
16:56
Unidentified speaker: And this one's also very close to my heart because I uh I helped get it going.
17:02
Unidentified speaker: And it's just a great thing to actually see this cross the finish line with all the efforts and work that so many people put into this.
17:10
Unidentified speaker: So talk about our new interoperability ISO standard.
17:14
Unidentified speaker: So, Aaron, as you mentioned um some years ago, and I confess that I can't actually remember how many it was, might have been 2018, might have been 2019.
17:22
Unidentified speaker: I'm pretty sure it was before COVID, that you actually took the stage at what we then held as the AMR conference.
17:30
Unidentified speaker: So A3 uh held an AMR conference, and you got up on the stage and you said that one of the biggest things that the AMR community and the AMR industry needed to do was to make sure that these machines could talk to each other.
17:47
Unidentified speaker: Because, you know, you were experiencing and you weren't the only one that uh these, you know, two machines that couldn't talk to each other because maybe they were from different vendors, would approach each other, block each other, and then they would engage in this standoff.
18:01
Unidentified speaker: And they, you know, that impeded their productivity, which basically impeded the value of the system.
18:07
Unidentified speaker: So um, so you said, listen, industry, you have to figure out how to get these things to talk to each other.
18:14
Unidentified speaker: Because otherwise, what happens is these two machines are having their standoff.
18:18
Unidentified speaker: A human has to go figure it out and turn one of them off, or I don't know what the human has to go solve the problem, and then they can resume their work.
18:27
Unidentified speaker: And so then that's just not a sustainable way to be.
18:31
Unidentified speaker: And so we took that to heart.
18:33
Unidentified speaker: We actually did.
18:34
Unidentified speaker: We we heard you so we heard you, and actually, um, I want to uh give a lot of credit to uh Mass Robotics who really took that and ran with it and developed a specification for North America, and so um, you know, we approached them and and we said, How would you like for this to be the seed document of a new ISO standard?
18:58
Unidentified speaker: And they could see the value of having it be ISO and having it be a more um, you know, the international versus only North America.
19:08
Unidentified speaker: And we also recognize that there were other uh regional or national standards that similarly are starting to approach this.
19:16
Unidentified speaker: And really to get to be the most useful, it has to be international because if you have an interoperability standard that is national, only one country uses it, you don't really have an interoperability standard.
19:28
Unidentified speaker: And so we took it to ISO, we invited the uh related uh national and regional um standards to come and take part and and and um contribute the the best pieces of their standard so that the ISO, the international version, truly would be you know what industry needs internationally.
19:49
Unidentified speaker: And so the end result was this um 21423, which is the new code for this interoperability standard for industrial mobile robots.
19:57
Unidentified speaker: And um it's quite a lot different from the original Mass Robotics standard.
20:02
Unidentified speaker: It's it's um it has had the ISO treatment, which means you know, people from around the world, many different countries have taken a crack at improving it.
20:10
Unidentified speaker: And and improve by improving it, meaning it will work for their industry.
20:17
Unidentified speaker: And so um, and that's really the that's really the value of standards is that um the wider the network, the the wider the network that uses them, the more valuable they are.
20:28
Unidentified speaker: And so um that's what we really hoped to achieve with 21423.
20:33
Unidentified speaker: And, you know, we do have a good working relationship with, for example, VDA 5050 people.
20:38
Unidentified speaker: In fact, we're working right now on a TR to supplement this 21423 that will help explain how um these two standards can work together in the same facility.
20:49
Unidentified speaker: So we're not, you know, this 21423 doesn't necessarily supplant or supersede anything that people have become used to using, such as VDA 5050.
20:59
Unidentified speaker: Um they can work together for different um modalities.
21:03
Unidentified speaker: And so, you know, we really tried hard.
21:06
Unidentified speaker: We really tried hard, Aaron, to get industry what it needed in the way of interoperability for mobile, mobile robots.
21:13
Unidentified speaker: And of course, no standard is perfect.
21:15
Unidentified speaker: I'll be the first to admit that.
21:17
Unidentified speaker: And so, you know, there's always opportunities to improve it.
21:20
Unidentified speaker: Yeah.
21:21
Unidentified speaker: And there is another ISO effort underway that's really special and close to, I know a lot of our members, and it's a first of its kind.
21:31
Unidentified speaker: I know we did some work on the pre-standardization research on it for humanoid stability, but we're not calling it humanoid stability.
21:40
Unidentified speaker: Yeah, we're having to use a very special term here.
21:43
Unidentified speaker: So let's talk about this.
21:45
Unidentified speaker: This new this other standard is still in development and you guys are still working on it.
21:50
Unidentified speaker: Give us an update on on where that standard is right now.
21:53
Unidentified speaker: Yeah, so that's um another one of the uh the growing number of ISO standards that have been um the projects have been started in recent years.
22:02
Unidentified speaker: And this is um 25785.
22:05
Unidentified speaker: So this is gonna be its number when it gets published.
22:07
Unidentified speaker: And it's it's already known by that number, 25785.
22:10
Unidentified speaker: The particular type of machine we're not calling a humanoid, so we're calling it um the dynamically stable industrial mobile robot, so or DSIMR.
22:22
Unidentified speaker: And so the IMR is the what we're calling that the industrial mobile robot, it's an industrial mobile robot.
22:30
Unidentified speaker: It's a mobile robot used in an industrial setting.
22:34
Unidentified speaker: The dynamically stable ones are unique in the sense that they require active management of their actuators in order to maintain their stability.
22:43
Unidentified speaker: In the absence of power, they will fall over.
22:45
Unidentified speaker: They will fall over, and that will create a new hazard.
22:48
Unidentified speaker: So um, you know, if you're if you have a mobile robot that's carrying a tote that is full of ball bearings and it falls over, and those ball bearings go all over the place, that's creates a new hazard.
22:59
Unidentified speaker: Even aside from, you know, if the robot falls on a person, that's a new a new hazard.
23:04
Unidentified speaker: And so um, and the the complexity of of these machines, um the complexity of the controls that they have to exert over their actuators to maintain their stability is um really quite impressive.
23:17
Unidentified speaker: And so it's really one of the better uses for for AI, in my opinion, is uh the use of AI to manage this for the robot.
23:27
Unidentified speaker: And so um but what it means is that what happens when you push the e stop has to be different for this class of machine.
23:37
Unidentified speaker: So when you push the e stop for an industrial robot or a statically stable mobile robot, um you push the e stop, it removes all powers to the actuators.
23:47
Unidentified speaker: But in the case of the dynamically stable, you remove all power to the actuators, you are not reducing the hazards to human, you're actually creating a new hazard.
23:56
Unidentified speaker: And so um, you know, this class of machines is uh we are we are creating a whole new standard, safety standard.
24:05
Unidentified speaker: Obviously, they are a specialized case of an industrial mobile robot.
24:09
Unidentified speaker: And so um, you know, we hope to be consistent with other principles of safety of keeping people safe around mobile robots.
24:18
Unidentified speaker: Um, but we additionally have this whole stability aspect to it.
24:23
Unidentified speaker: It includes humanoids, so dynamically stable DSIMR includes humanoids or bipeds, but also any other dynamically stable form factor.
24:33
Unidentified speaker: So it also includes quadrupeds, it also includes wheel balancing robots and potentially other form factors that we haven't even thought of yet, as long as they require active management to remain stable.
24:46
Unidentified speaker: And so we're really excited about that.
24:48
Unidentified speaker: We started that project about a year and a half ago, and like I said, we're about halfway through the three-year development cycle.
24:55
Unidentified speaker: And in fact, we're meeting next week.
24:57
Unidentified speaker: Next week, this group is meeting in Stuttgart, Germany, and uh hoping to continue making progress on resolving our CD comments.
25:06
Unidentified speaker: Uh, hopefully we'll we'll get to our DIS ballot version in May of 2027.
25:13
Unidentified speaker: Um and so we're just continuing to discuss the comments and decide which ones we want to adopt.
25:21
Unidentified speaker: Is there anything else that's on the horizon that you'd like our audience members to know about and how they can even maybe get involved in some of these?
25:29
Unidentified speaker: Yeah.
25:29
Unidentified speaker: So one of the things that I mentioned um for these for the use in DSIMR safety is um AI or machine learning, because the complexity of managing the stability of these machines is such that you know you can actually never have a true manual mode.
25:46
Unidentified speaker: It's not actually possible for a human to make all of the tiny adjustments needed to all of the actuators to keep the thing stable.
25:54
Unidentified speaker: And so um basically uh, you know, AI is starting to come into its own in some of these areas that are invisible to people because you know it's it's internal to the dynamically stable robot.
26:07
Unidentified speaker: And so the other thing about AI is you know it's it's going into new places.
26:14
Unidentified speaker: And so one of the things that we're we're actually watching very carefully in the DSIMR standard is um the development of and the use of a new paradigm about AI and functional safety.
26:28
Unidentified speaker: And that's taking place in a different technical committee.
26:31
Unidentified speaker: Um actually, it's a joint effort between IEC and uh and and ISO.
26:36
Unidentified speaker: So ISO and the IEC are two different international standards bodies with that focus on slightly different uh aspects of machinery.
26:46
Unidentified speaker: Um but in in some things they have a joint, they have a joint technical committee, um JTC1, joint technical committee, and uh this artificial intelligence standardization is one of those things that's being tackled jointly between ISO and IEC.
27:02
Unidentified speaker: And so there's actually a new standard being developed now on um how AI impacts functional safety or how to use AI in functional safety or how to adapt the concepts of functional safety for an AI world.
27:18
Unidentified speaker: And so um we are anticipating that will be a very important standard uh that the DSIMR standard will normatively reference.
27:27
Unidentified speaker: And um, so we're watching that pretty closely.
27:30
Unidentified speaker: And that's actually one of the things we have a we have the convener, or that's the special standards word for chair, the chair of the working group in JTC1 that's developing that standard on AI and functional safety, is actually uh he's gonna come and give us a presentation about where that work stands.
27:49
Unidentified speaker: And so I'm really excited to hear about that.
27:52
Unidentified speaker: And uh, and we're just watching that very closely.
27:55
Unidentified speaker: Darryl, it's always great to have you come on the show and talk to us about standards.
28:00
Unidentified speaker: It's not always the enthusiasm or all of that of that gets folks going in robotics, but it's the so important to why we actually have safe robotic systems out there.
28:14
Unidentified speaker: And I highly encourage if this is a topic you in the audience are interested in, reach out to Carol here at A3, get involved.
28:23
Unidentified speaker: She her and her team would gladly have you on board on any of these efforts that she just talked about today.
28:30
Unidentified speaker: Uh so Carol, again, thank you so much for everything you and your team do at A3 and all the work you're bringing uh to make everything safe for us across the board.
28:40
Unidentified speaker: And now we have an interoperability standard I'm so excited about.
28:44
Unidentified speaker: So, Carol, anytime you want to come back on the show, you are always welcome here.
28:48
Unidentified speaker: And for that, folks, that is our show this week on the business of automation.
28:52
Unidentified speaker: I'm Aaron Frather, Director of Market Intelligence here at A3, the Association for Advancing Automation, and we'll see you next week.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3 and interpreted by Qwen3.6-35B-A3B-GGUF:UD-Q4_K_XL. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.
All insights