Thought Leaders in Artificial Intelligence: Darvis CEO Jan-Philipp Mohr (Part 2)

Sramana Mitra: What does the competitive landscape look like?

Jan-Philipp Mohr: There are a lot of AI companies and computer vision companies out there, but we try to concentrate on the logistical aspects of it. There are not that many competitors. On the one hand, there is RFID. On the other hand, there are some US companies and British companies who are doing computer vision in a similar context. There’re some more on the retail side and some more on the strict detection element. We tried to concentrate on the mapping aspect of things.

The combination of this geolocation comes out of the box with our platform where we create a lot of proprietary products and software. We don’t need to train models to find positions and things like that. The customer can turn any computer vision model into a model that detects position and other things in relation to a map which is a big advantage.

When you think about geolocation and its process, the biggest impact for credit card companies in the past against fraud was the location. If you’re usually spending money in a certain area and you haven’t bought a flight ticket, you are in a city where you’ve never been and it’s not your usual pattern, they would automatically decline that transaction.

Geolocation has a huge impact on processes today. A lot of the other solutions require you to enter data manually. We generate that data out of the real world, so location is an important aspect in our value proposition – understanding where in the process something is.

Our solution is down to a centimeter in accuracy. An RFID, for example, is between a meter and more. We’re a lot more accurate and we don’t require batteries. A lot of active RFID chips require batteries. We do the tracking for the operating rooms for the tools to understand unused items. Position is very relevant. Where are the tools? It’s not enough to just take the tools. It’s essential to detect where the tools are and whether they have been used already. You need to be very accurate.

Sramana Mitra: In your US Postal Service use case, if something is sitting on a warehouse shelf for a long time, that creates an alarm.

Jan-Philipp Mohr: More or less. It’s more about the assets around that.

Sramana Mitra: Are you also inside the vehicles?

Jan-Philipp Mohr: The nice thing about technology is, we can be inside vehicles. We’re not yet in the electric vehicles they just acquired, but the technology is ready for that. Once they deploy it in the warehouse and train the models, they can easily retrain it in the vehicles.

Sramana Mitra: How do you calculate ROI?

Jan-Philipp Mohr: Very straightforward. Dwell time is a very big cross factor. The more I can decrease dwell time, the more I increase flow. When you talk about wall-to-wall visibility, the quicker something moves, the more dollars I generate.

Sramana Mitra: You’re talking about your retail and logistics use case.

Jan-Philipp Mohr: Yes. In hospitals, it’s different. We are deployed in both German and US hospitals now. Germany is well-known for privacy, so we had to go through a lot of hoops to create a privacy-compliant solution. There is no screen-to-camera principle in any of our solutions. You don’t see what the camera sees. You only see the metadata. That puts up the requirement to produce high-value data. We learned that you can never look at 40 cameras at the same time. There’s no value in looking at the cameras in reality; there’s only value in generating data with relevant insights for your process.

Sramana Mitra: If the nurse needs to go, that’s an actionable alert.

Jan-Philipp Mohr: Yes, it’s all about taking out as many boundaries for the technology as possible so that somebody can put his process into the platform and then understand how to improve and how to measure, and take that measurement without any touch. The iPhone got big because of the touchscreen. We’re in that next generation of possibilities where applications can, without touch, generate data and improve your environment.

This segment is part 2 in the series : Thought Leaders in Artificial Intelligence: Darvis CEO Jan-Philipp Mohr
1 2 3