cover of episode EP 576: Sientis is Reinventing Inventory Accuracy with Drones

EP 576: Sientis is Reinventing Inventory Accuracy with Drones

2025/4/2
logo of podcast The New Warehouse Podcast

The New Warehouse Podcast

AI Deep Dive AI Chapters Transcript
People
J
Jim Mass
M
Mouhyemen Khan
R
Rajiv J Gandhi
Topics
Jim Mass: 我们Sientis公司(原Nokia Ames)的无人机库存监控服务(AIMS)旨在解决仓库库存盘点效率低、准确率差的问题。通过在仓库中部署配备先进人工智能和计算机视觉技术的无人机,我们可以自动扫描条形码、二维码和车牌号等信息,并将数据传输到用户界面。这将帮助仓库运营商识别丢失的库存、错误的库存和不匹配的物品,从而提高库存准确性和运营效率。无人机体积小巧,适合各种规模的仓库,可以显著减少人工盘点所需的人力,降低成本,并让员工专注于其他更重要的任务,例如拣货、包装和订单管理。最终,我们的目标是让仓库经理能够随时随地轻松查看仓库的库存状况,从而减轻他们的负担。 Rajiv J Gandhi: 我们的用户界面提供每日扫描数据,包括扫描日期、通道、位置数量以及WMS数据分析结果。用户可以查看丢失库存、找到的物品、错放的物品以及WMS中未记录的物品等信息。系统还会提供相应的图像信息,并根据需要对图像进行分类和筛选,例如突出显示最重要的图像。此外,系统还支持用户取消审核并重新检查,方便沟通和协作。我们的系统不断更新迭代,提供更详细的库存信息,例如货位级别的信息和产品信息,帮助用户优化库存管理。系统能够识别整托盘的货物,无需逐个计数,提高效率。 Mouhyemen Khan: 我们的无人机解决方案采用自主导航技术,无需在仓库安装任何外部标记,能够快速部署并适应各种仓库环境。无人机体积小巧,能够在狭窄的通道中飞行,并具有较长的续航时间(25-30分钟)和高分辨率摄像头,能够进行精确导航和数据采集。无人机能够扫描各种类型的条形码和字符,并进行箱子计数,提供多种库存数据分析。我们不仅提供图像数据,还进行数据处理,为客户提供有意义的分析结果,例如丢失库存、缺失箱子等信息。在单次充电下,无人机可以扫描数百个货位,多次任务后可以扫描数千个货位,效率远高于人工盘点。

Deep Dive

Chapters
Sientis, a Nokia venture, utilizes drone technology to improve warehouse inventory accuracy and efficiency. Their Autonomous Inventory Monitoring Service (AIMS), now rebranded as Sientis, addresses the inefficiencies of manual counting methods, offering a significant ROI by reducing errors and freeing up employees for other tasks. The drones are designed to fit various warehouse sizes and significantly reduce the number of employees needed for inventory.
  • Sientis uses drones with AI and computer vision to improve inventory accuracy.
  • Manual inventory counting is time-consuming and error-prone.
  • Sientis' solution offers a positive ROI by reducing labor costs and increasing efficiency.
  • Drones can fit into small and large warehouse spaces.

Shownotes Transcript

Translations:
中文

You don't want your employees to be idle, so why would you want your robots to be idle? Zebra Technologies is addressing idle robots by bringing their latest solution to the market, which requires 30% fewer robots.

resulting in increased asset utilization and a cost-effective entry to automation for your fulfillment operations. The Zebra Symmetry Fulfillment Ecosystem is bringing you balanced utilization, more cubic capacity, and optimized pick paths so you can operate at peak performance. Want to take your fulfillment operation to the next level? Learn more at zebra.com or click the link in the show notes.

The New Warehouse podcast hosted by Kevin Lawton is your source for insights and ideas from the distribution, transportation, and logistics industry. A new episode every Monday morning brings you the latest from industry experts and thought leaders. And now, here's Kevin.

Hey guys, it's Kevin with the New Warehouse Podcast and I am here at Manifest 2025 and right now I am in the Nokia Ames booth and we're going to talk to Jim Mass of Nokia here. He's going to tell us a little bit about what they're doing, what they're all about, what the latest is since the last time we spoke to them at Modex and then we're going to get a little demo of the drones themselves. So Jim, how are you? Absolutely doing great, Kevin. Appreciate you having me.

Today, to talk a little bit about Nokia Ames. I know you mentioned last time we connected was at MODEX, so a lot of improvements have occurred since then. So a little bit about Nokia Ames. Ames stands for Atomic Inventory Monitoring Service. We're out there to solve the goal to make inventory and cycle counting much more efficient, more accurate, and be able to provide warehouse operators with a positive ROI. And by doing that is by deploying drones throughout the warehouse, and it's going out there pulling anything from the barcodes, QR codes,

license plate numbers and then bringing it back to us in our URI user interface which is going to give them the ability to see anything from lost inventory, mis-inventory, mismatched items. It's all about making it

simple and more relaxing for the warehouse operators, warehouse operators, inventory individuals out there today. All right, great. And we're going to see that drone in a minute here and how that works. But tell us a little bit about the problem you're solving first, right? Let's understand, how big is this problem for operations?

It's huge right now. If you think about now, a lot of operators right there are going about manually right now. They're going out there in scissor lifts. They're going out there with clipboards, pen and paper, going with scanners out there. It's time consuming. Errors occur. Our customers right now that we actually implemented with...

are having issues, right? Employees aren't showing up, right? They're making errors on a regular basis. And that type of stuff adds up. And we're out there to solve that, to make it more simplified, to let those individuals and employees that are going out there doing inventory counting, going out there in front of the warehouse, and having them more focused on more important jobs and tasks, pick and pack, order management, stuff like that. So we're making everything much more simplified, like I think about efficient,

Much more efficient, much more accurate, and we give that ROI in the long run. Yeah, absolutely. I think that the accuracy is such a key for operations because it not only helps you maintain a better customer experience at the end of the day, but then overall the operation just runs smoother, right? With a lot less speed bumps and a lot less hiccups along the way when you have inaccurate locations or whatever the case may be. Absolutely. We want to make sure that the inventory manager, operations manager can be sitting at his house, can be sitting on the beach,

worry-free, right? Can hop onto our UI, pull it up and be able to see exactly what's happening in this warehouse at any time. Yeah, yeah. I think that's fantastic. And tell us about the, I guess, the decision to utilize a drone as the vehicle for the data collection. Yeah, so I think drone, it's much more smaller. You kind of look out there at the competition, you see some bigger robotics out there that's doing inventory management, much more simpler. Our drones can fit in smaller aisle spaces.

So we got the right size, the right fit for small, medium, large companies out there. From warehouses from 100,000 square feet all the way up to a million square feet. - All right, interesting, interesting. And from a drone perspective, I mean, how does that kind of augment the labor? Like you look at your traditional inventory team within the warehouse, you have some cycle counters, maybe you have like a lead that's doing some problem solving. What does a team on inventory control look like after you implement the drones?

So, yeah, so we typically talk, we talked to someone today that have 12 full-time employees that are going out to do an inventory. So that's going to cut that significantly, right? Right. I think the cost when it comes to having an employee on site is costly, right? Yeah. Like I talked about before, we want to take those employees and go focus on more important tasks within the warehouse and let our drone do the work for them when it comes to inventory. All right.

- All right, awesome, definitely. - Yeah. - And anything else we should know about the drone before we see it kind of fly up here in the air? - No, I think, you know, in the two years we've been doing this, we have made significant strides. Improvements are happening month after month. This team that we see right here today, they're amazing. Some of the smartest individuals I've ever been around.

It's impressive to see what they do here at Bell Labs. All right, so we just talked to Jim about Nokia and the Ames solution that they have here with the drones. But we're going to check out the inventory platform that they have here to be able to show what's going on. And Rajiv is going to show us that. So walk us through a little bit about what's here on screen. Yeah, absolutely. What you see here is the daily scans of what the drone sees.

And from there you get to know a few things, everything from the data of course, the aisles we scanned, the amount of locations that we scanned that night or the day, and your ability to know that our examination of everything that came from the WMS to our analytics is done and that your ability to view that data.

Interesting. I see like 1,200 scans here. That's a lot. Yeah, it's a lot, right? And that's all within maybe like a half an hour scan. Oh, wow. That's pretty good. So, you can imagine through an eight-hour shift what you can get out of that. Yeah, absolutely. Absolutely. But in this case, just as a micro sample, we'll go through the first one there. So, when you enter the UI,

the ability to sit anywhere you want, whether it's in your office, at home, or otherwise, can examine the data here. What you tend to see on the other side of this is that even as a warehouse manager, you get to see that someone has actually reviewed the location already.

Right, through the UI. So you know that people on the floor are looking at this and that they're examining the data. To that point, we have categorized a lot of the analytics into separate pieces such as lost inventory, found, misplaced. We even notified things that are not in the WMS that we found as well, as well as unable to identify, which lovingly lets you know as a person on the floor where there might be plastic hanging over. Oh, okay. Right, that you need to clean up that aisle.

that a pallet is sticking out, right? Or that really that the box is not facing outward in a particular way that it's not facing forward. So, you know, things that help you manage the warehouse efficiently. Yeah. So it's not just about counting. No, it's a lot more than that. And I think what we try to do with this is give it within a three persona realm, right? One is the person who's running the warehouse, who's like a service center manager. Okay. To a guy who's managing 12 people on the floor. Yeah. To the person who's actually going out and counting.

Right? So, we try to satisfy those three personas here. And that helps really have a cohesive environment all through the UI. Yeah, absolutely. I can see that. So, even if I go to such a thing as misplaced, and I think what's good to note is, we'll definitely tell the user, not only is it reviewed in that location, but what we saw, what should be there, the count of what we saw, and what the WMS says should be the count for it. And if you wanted to,

that person who's using it, the UI can simply examine further. And what you get to see is the actual images of what the drone saw. And through our technology, we are able to not only give you those images, but choose out of the

30 some odd images we took of that space, identify the four to five very best images through our technology and front load them there. Not only showing what we saw, which is the barcode, but also the bin boundary of what we saw. So, we virtually put that on there to show where the drone is really examining data. Gotcha. So, it's kind of identifying like- The space in which that is, right? Almost the boundary of that, right? Interesting. And to that point is what's really good about the data that we collect is, we'll even let you know

If let's say something is in an adjacent bin and it really should be in here, we'll even tell you it's a match. But guess what? It's just not in that boundary. Got it. So, you'll get to, you know, so we don't need nitpicking about it, we let them know the full picture of that world there. And it allows you to really see not only navigating from one picture to the other to see where the drone saw, your ability to zoom in and out to a very high fine degree with the 4K images that we have, and really easy navigation.

Now, as a manager, you might be looking at this. At the same time, a person in cycle counting might see this. But the great thing is even though someone's reviewed it and you find a discrepancy, the manager can say, guess what? I want to unreview this. Please go check it again. Really? Okay. Right. So there's a good communication loop happening within the UI as well. Yeah. And noticing that for every night. Very interesting. Very intuitive, too. And I think it allows you to take away some of that.

you know, manual communication, right? Absolutely. Where you have to go find somebody, oh, go check it again, right? And then maybe they forget, right? But, you know, this kind of dictates, like, the work a little bit, right? Absolutely. And, you know, part of the innovations is we release very frequently. We have that just very...

as the word says, agile process. But we release so frequently that we're giving you bin level information really, but we also want to dive down into the product information as well. Interesting. Right? So whereas a product might be in one bin and is supposed to be in five others, six others, and then you can rectify those situations, know exactly where to look,

and we're to avoid locations as well. Absolutely. Interesting. And I noticed here, this line here, I'm curious about this. It says that the AMS count was fully wrapped. Yes, yes. What does that mean? So part of our technology is saying, you might want to count the case counts in the bin, in the location. But a lot of times what people want to know is,

Rather than that, do we see a fully wrapped pallet? Is it supposed to be there, right? Right. That particular item, you don't need to know the case count because it's fully wrapped. So, we already know everything's okay there. Got you. Okay. Very interesting. So, it's recognizing that there's just full pallet. Yeah. The technology is really good in the context that we identify the fully wrapped pallet. We tell you what we can't count. We can tell you that there is an empty bin as well. Right. So, we try to identify those and we're working every day to not only identify that the case counts, but maybe you have reels. Yeah.

and you want to be able to count those reels. So starting to use those AI technology to identify particular types of products as well in the future. Interesting, very interesting. Well, we appreciate you showing us the UI here, Rajiv. Thank you, Kevin. We're looking forward to now seeing that drone in action. All right, so now I'm going to be talking with Muhammad here, and he's going to talk us through exactly how the drone, as we see it just took off, exactly what the drone's doing and how it actually works within an operation.

So tell us a little bit about what it's doing here. Absolutely. Thank you for having me, Kevin. So this is the Nokia Ames drone solution that is powering our indoor autonomy in warehouses. We actually have built the software by ourselves fully using their software is proprietary. We're using advanced AI and computer vision algorithms to actually autonomously fly the drone in our customer sites. As you can see that

In this cage, we don't have any external markers or anything installed and we do not do any kind of installation even in the customer site. The drone is fully capable of flying autonomously in the customer facility. So, this actually enables us to very easily and quickly deploy in customer sites. It's flying autonomously, it's a small platform so it can actually even fly in narrow aisles which is a huge advantage that we have over some of our competitors.

So when there are narrow spaces between the towers, the drone is actually able to fly between them quite easily. We have great battery

time on these drones upwards of 25 minutes, sometimes even up to 30 minutes. So, we are able to cover quite a bit of distance on a single charge. We have high resolution cameras on the drone which actually allows us to do very precise navigation as well as capturing high resolution analytics. And those images as you have seen in the UI where we actually had renderings, but our customers actually get high definition images coming out of these drones.

Interesting. So, on the locations, right? So, basically if a warehouse already has their own location barcodes, it just recognizes those? There's no special location naming or anything that needs to happen? Right. So, we are able to actually do any kind of barcode scanning. Okay. We are also able to do optical character recognition. So, if there are characters on the boxes, we can actually scan those as well. Not just that, we also do box counting.

And we have even more use cases actually planned for the year. Some really exciting things are coming out later this year, providing very high value data to our customers. So it's not like we just capture the images and give it off to the customer. We are actually doing a lot of engineering processing on our end and giving them very meaningful analytics that they can actually take a look

over they can actually look back and see where maybe there's lost inventory maybe there's a box missing not just that even how many boxes or cases there are in that particular location okay right so we are able to do all of these things and that's what we actually provide to the customers today so it goes even beyond just barcodes oh wow okay very interesting and and now on the the battery side of things you said that the battery can last for 25 maybe 30 minutes right so for perspective

in that time frame, how many locations can a drone count? Right. So that is both a function of how tall the tower is and how wide the tower is also. So we are uniformly scanning the towers. We also can actually have custom missions designed for the customer where they want to actually scan specific locations.

And so that way it depends whether you want to do wall-to-wall scan, then we will actually uniformly scan everything and it's significantly faster than what a human would actually do because a human would have to get on a forklift, go all the way up, scan something and if they have to just go to the adjacent bin, they have to come down and then go back up again. But the drone is just nothing but a rightward movement.

So you're saving a ton of time. So the number of bins that we can actually scan on a single charge is high hundreds. And that depends on the density of how many bins there are, how tall the tower also is. But by the end of the mission, by the end of the night where we actually have run multiple missions, we're talking about thousands and thousands of bins scanned by just a single drone. Yeah. Yeah. I saw in the UI there, there's about 1,200 or so scanned, right? Absolutely. In one day. Yeah.

So very interesting there. So basically you said even though the battery is lasting just 25-30 minutes, you're still able to get a lot of scans within that time frame. Because you're not wasting any time to go to the neighboring bin. You're doing the necessary movement required just to go to that bin. A human would have to load, unload, go up vertically, come down, move to the adjacent bin, so on and so forth.

we are removing all of those inefficiencies that exist in the human cycle operation. Interesting. And it sounds like deployment of this is super simple as well. So, tell us a little bit about what is that?

What does that look like? Right. So, on the deployment strategy, we do site surveying of the customer facility. The floor plans are used, our engineers actually do all the calculations necessary to design sequences of missions. And very quickly, we're able to generate large numbers of missions depending on the size of the customer facility.

The drone is then able to actually go and execute those missions. The customer will be able to place the drones in specified checkpoints where the drone will take off from, scan those locations designated by the missions, come back, return, and then they can execute the next mission after that.

Interesting, interesting. So very interesting to learn about the Nokia AIMS solution here and see the drone actually in action and then what it can produce on the UI itself. So really appreciate you and the Nokia AIMS team for having me in the booth here to show us the solution.

Pleasure to talk to you, Kevin. Thank you very much for having me. Definitely, yeah. And if people want to find out more information, what's the best way? You can reach out to us on our website, our LinkedIn profiles, as well as our email addresses. Watch out for that. Our website is coming out soon. All right, there we go. And we'll put that information in the show notes and at thenewwarehouse.com as well and let us know for sure what do you think about drones doing your inventory counting.

You've been listening to the New Warehouse Podcast with Kevin Lawton. Subscribe and check us out online at thenewwarehouse.com. Enjoyed this episode? Make sure you are subscribed to the podcast and for more content from The New Warehouse, find us on LinkedIn and YouTube. Links to subscribe can be found in the show notes and for everything The New Warehouse, head to thenewwarehouse.com.