In new safety guidance on autonomous systems, the Global Mining Guidelines Group urges companies to adopt a comprehensive change-management approach. Potential procedural conflicts between manned and automated operations need to be addressed for technology to stay safe. In a recent interview discussing strategies for navigating this shift to an integrated safety ecosystem, Neville Judd, Hexagon Mining Communications Director, spoke to Craig Ross of Ross SH Consulting and Hexagon Mining Vice President, Autonomous, Andrew Crose.

NJ: Craig, you have 30-plus years of experience in the industry. For seven years, you were VP of Safety and Health at Barrick. You spent much of that time working to improve safety. What procedural conflicts could you see arising between manned and automated operations that need to be addressed for technology to stay safe?

CR: At Barrick, we were pursuing the autonomous venture quite strongly. And one of the things that I noticed that started to bother me in the beginning is the desire to immediately jump from what we were doing in manned operations to jump completely over to autonomous operations and really forcing the issue. And the problem that I had with it is the general risk-management approach was to isolate the autonomous fleet or operations from any people, which would require that any human interaction with the autonomous fleet would require complete shutdown of the autonomous operation. And to me, that wasn’t feasible, because when you start really looking at the operation and the things that need to occur in conjunction with the operating practice, the risk was being managed by isolation when we should be looking at how we manage the risk with humans around, which is much more difficult and one of the reasons I liked Hexagon’s approach. But some of the conflicts, for instance include pre-shift examinations. You’ve got to inspect your equipment so that requires complete shutdown of everything. Whereas, the goal of autonomous is to be able to continue running. Workplace exams – you’ve got to get in there and take a look at your work areas legally, typically on every shift basis. Other things – we depend on humans to give us feedback on operating conditions, hazardous conditions such as the conditions of waste dumps, if there’s tension, cracked berms, highwall activity, things of that nature. I worry that we’d lose some of that. To me, that’s a bit of a conflict. It has to be resolved. Road maintenance. In general, you know, when you get spillage or deteriorating roads, typically you’re going to need RTDs or primarily graders, which are typically a human situation entered into that autonomous zone, and if we’ve isolated the zone, we can’t do that, so we have to address that. Even out on the dumps, where the trucks are dumping, typically you have a dozer and RTD out there maintaining those dumps, pushing the material over, making sure the area is safe for the operators. But now, do you lose that? I mean, you can’t operate without that. And that seems, to me, is just some of the conflict. But to me, the real answer is to when you do the comprehensive risk assessment, rather than try to manage the risk of humans being around the equipment by isolation, primarily isolation, I think that we need to manage the risk with the humans being present, similar to what they’re doing with autonomous vehicles that we’re starting to see in our cities and in towns, that you’ve got to manage the risk associated with human interaction.

NJ: Let’s talk about a specific example. Andrew, I know in your work with Hexagon in South Africa, you’ve worked closely with Anglo American, who are going beyond just compliance with the Zero Harm initiative. How are they using safety technologies to manage risk and instill organisational change?

AC:  So, across the Anglo portfolio of mines, we’re working with many of them on different technologies and different implementations. It’s nice that they take a very holistic approach to it, so across maybe sites we’re working with them on their fatigue management and driver-distraction program. And in some regards, these have been some of the initial implementations we’ve done in our OAS fatigue program. We’re finding the algorithm, improving the system, and making it more user friendly. And the same, we’ve also developed alongside their mines and others are our collision-avoidance system all the way through the auto-braking. And it’s interesting going on the journey with them, because they’re, you know, they’re a very innovative group and have a lot of great ideas they bring to the table. They look at the system holistically. I mean, not just the technology, but the change management, the implementation, how it gets accepted by the operations, how it gets accepted by the unions and the communities. It’s been a real good journey for us working with them.

NJ:  I’m glad you mentioned change management. How does change management differ when autonomous systems are introduced?

CR:  I don’t think there’s too much difference in how, if you’re doing change management the proper way. I think, the difference with autonomous is you’ve added a whole new activity to the situation. And I think you need to really line out all the tasks that need to be performed throughout the operation and then start introducing the autonomous and what hazards it’s going to present. And again, as I mentioned earlier, the worst-case scenario is that we just jump immediately to isolation. Not saying isolation can’t be the answer in some cases, but I think it’s a mistake to jump right to isolation for the autonomous mining. I would like to see us when we identify those risks in our change-management process is, okay, how can we have the technology help us manage that risk with those human beings present and not just strictly rely on isolation? I think it can be done. And it’s one of the things that I learned when we were working on the collision-avoidance technology is the information that we were getting from the data by implementing the collision avoidance gave us so much more information that we didn’t have before on what’s really going on out there, because the near misses or close calls wouldn’t resonate with the human being, but you would see it on the data that we were getting back from the collision-avoidance technology. And it really helped us to see what was really going on out there and would help us better manage that risk. And so, the technology, I was getting really excited on how much we were learning about the risks and better ways to manage it.

AC: One of the other things that I think has been interesting in the journey is also the engagement of the local community. So, we recently, in the last year or so, had been working with a firm in the little community near Kathu in South Africa to help support the system coming by the name of DeNovo. And they’ve been great in assisting us on the maintenance of the program locally to still have an engagement of the local community, because that’s always, obviously, a big fear when it comes to autonomous is that you’re removing jobs from local community. And I think in our engagement, we’ve actually shown that we’ve increased the highly skilled jobs in the local community much more so.

CR: My approach to that is similar to what they did in the airline industry with all of the accidents or disasters they were having back in the ‘80s, is the more technology they were putting into the aircraft, it wasn’t to necessarily to replace pilots; it was to help pilots not make errors. Before they really took a strong look at it, they blamed everything on pilot error. And it isn’t like I haven’t seen that in the mining industry. Everything’s operator error. But everybody makes mistakes. So how can we implement the technology to help ensure that our operators don’t make mistakes or they get a warning if a mistake is about to be made, and initially that’s the biggest benefit I saw from it was to help our operators be better and safer operators in the long run.

NJ: Andrew, what are some of the ways that Hexagon is approaching safety and autonomy from the perspectives of both technology and change management?

AC:  A couple of things that we’re doing, one is taking it from a building-blocks approach. When we approach a new project, it’s not that we just go in necessarily in a big-bang approach to try and do a complete step change. But instead, we’ll look at the entire risk profile, as Craig mentioned, and do an assessment of what is the best strategy to automate portions of the process or portions of the mining sequence or possibly and greenfield the entire mine, instead of trying to approach it as one big change. And I think this helps keep the fear of change somewhat at bay because they see it as a continuous process. Again, with local engagement, mine engagement, operator engagement, and then, obviously, supported by the data, then we can definitely see that benefit to the operators and the users in the community. As well as we take a very high standard when it comes to project management. So, it’s very important to us that we approach our projects in a very professional way. We have very clear charters, very clear definitions, and a project plan that goes not just from kickoff meeting to handover, but kickoff meeting all the way through the ongoing program of running this system and programs like our Serious about Safety framework and others that help us maintain the gains that are envisioned in an autonomous project.

NJ:  Finally, what should technology companies be prioritising on their roadmaps when it comes to addressing the shift to an integrated and increasingly automated safety ecosystem?

CR:  I really agree with how Hexagon has been approaching this. And I really realised this early on when I started seeing the complexities of us just starting with the collision-avoidance technology. You know, there was a lot of issues we needed to work through and fix. And I started thinking about it holistically, and it goes back to rather than make this huge, giant step-change right over to autonomous, the step-by-step approach is going to be more beneficial in the end, because one of the examples or analogies I thought of is when we first saw automobiles show up, you know, back in the early 1900s, I mean, we didn’t all of a sudden say, “We have an automobile. Well, now let’s drive across the country with the family at 80 miles an hour,” because the infrastructure wasn’t in place in the vehicles, the technology wasn’t really available. But if you look at that over time, each new step along the way where they made improvements on infrastructure – you know, roads, gas stations, and improvements in the vehicles themselves—we’ve seen it advance over time and then even got to the point where instead of focus on how fast and far we can go is how much more safer we can make the vehicles, because as we started going faster, we started seeing more risk. And I think that the approach to this technology needs to be in the same manner. That’s again, why I liked how Hexagon was going about it, because they were focused on the collision avoidance, which was the first issue to address, and working through all the bugs to make sure that we trusted the technology, that it was seeing what we wanted it to see. It was alarming and responding when we wanted it to alarm, get rid of the false alarms, and working through that process. Because once you get it to the point where, okay, now, I completely trust it, from a collision-avoidance standpoint, now we can take the next step and move forward. Whereas, if you just do that huge step change all at once, there are so many things that you’re going to run into that you didn’t anticipate, even doing a great risk assessment. I know that from experience. There’s always something, whoops, I didn’t see that. There’s always something. So, if you take it step by step, you can capture those before it’s a catastrophic event. The other problem with trying to do it all at once is there’s so many issues to deal with you become overwhelmed, and then the people that resisted change, it just gives them more ammunition to resist that change because of all of the problems that are associated with it. So, if we can do it step by step, resolve each of those, and in time, people will trust the technology more and be willing to take the next step. And I think that’s one of the biggest issues with autonomous right now, why you see that isolation again is the primary risk-management approach is people don’t trust autonomous technology yet. But if we go about it right, and I think Hexagon is, we can get there, and it will take some time.

AC: I would say that one of the priorities that really needs to be on focus is understanding the technology and the implementation needs of the mine, and I think this is sometimes lost in the translation. So, mines have different focuses. Some might want to be, you know, really cutting edge and chasing that last bit of value with the newest and latest technology. That’s great for a company like Hexagon, that allows us to prove our technology. And the same other mines want to be implementing proven technology. And that’s also great because it allows us to deploy what we’ve learned in those early adopters. So not having that one-size-fits-all to all circumstances of a leading-edge startup, looking for anybody and everybody to try a system on or a value system that just wants to be the lowest-cost solution. But understanding what are the goals of the mine, what are they hoping to achieve in the safety automation, and then what’s the right approach to that situation within providing a top-tier safety ecosystem that you can stand behind and feel proud that you’re saving lives?