Flying Security Guard Startup Aptonomy Profile

Published Oct 27, 2016 10:49 AM

Backed by Silicon Valley's biggest incubator and led by two robotics PhDs, Aptonomy is set to launch, what they describe, as 'flying security guards'. [link no longer available]

Founder and CEO, Mihail Pivtoraiko [link no longer available] spoke with us about his company strategy for this report. We examine their offering, technical details, pricing, positioning and potential for integrators.

Drone ********

*** ******** ** **** ***** ****-****** **** ** ****** **********. ******** *** added *******, ******, * ******* *** microphone *** ***** ********** ** *** airframe ** ****** ***** "****** *****".

****** **** ** ~** *******, **** a *** ***** ** **** ***** (according ** ********, **** ****** *** out ****** *****), *** *** *** ** eye-level ***** ********* *********, ********* *** company **** ** **** ** ***** unique *********** *******. *** ****-********** ***** is ******** ** **** * *** speed ** *****

*** ***** *** ****-**** ** * charging ****, ***** ** ******* *** ~45 ******* ****** ***** ******* ******.

*** ********* ***** ***** ** ***** demo ** *** ***** ** *****, highlighting *** ******* ****** *** *******:

*** ******* ** * ******** ****:

*** "****** *****" *** ** ********** to ****** * ********, ** ****-******** to ** ***** ****, ***** ** operator ***** ** *********** **** ***** sensors, ** ******* ****** ************ ** an *****.

** ***** ********* *** **** ****** to ***** *** **** ******* ********* intruders *** ****** ****** ************** ** remote *********.

Company ********

*** *** **-********, *** ****** ********** [**** no ****** *********] *** *** ********* ***** [**** ** ****** *********] both ******** ***** ***'* **** *** Robotics ********* ** ******** *****. **** received * $***,*** **** ********** **** * **********, *** ** ******* ******'* **** selective **********, ** ****** ****.

****** ********* ** ********** ****** ********* round ** ** *********** ******, *** company ** ******* ** **** ********** developers, ** ******, *** * ** of *****.

********* **** **** * ***** ********, a *** **** ************ *******, ***** they *** ***** **** *******.

Deployment ********

******** ******* *** *** *****-***** ** be ** ******** ****** **** "**-****", because *** ****** **** ** ******* to ** ******* **** ********* * 4-drone ******* *******, ** **** *** constant-patrol **** *** ** ***. *** drones *** ******** *** **/* ******** (weather **********), *** ********* *** ****** to **** **** **** ****** ****** off-hours.

*** ****** *** ******** ** ** a ***** ********** *** **-**** ***** guards, ******* ** ****** ***** ***** *** awareness.

Optimal ***** *** ****** ******

********* **** ***** *******, ********** ******** layouts, ** **** ********** - ********* yards, ************ *****, **** **********, *** and *** **********, *********, ***., ***** make ******* ********** *********. ***** ***** of ********** **** ** **** * high **** ******, *** * *** density ** ******, ****** *** ***** and ****** ** * ******* ***** less ** ** *****.

*******

***** **** ** * $*,*** *** drone *****/******** ****** *** * $*,*** MSRP *** ***** ******* ***** ***. The ***** ****** ********, * ******* swap ***** ***** *****, *** * yearly **** ** *** ****** ********.

**** ***** *** *-***** *********** ***** (at ~$**,*** ****) ** ******* ********** in **** ** * ****** ********** guard *** * ***** *** ***.

******** ** ******* ** **** ******* integrators *** **** ***** ********* ** those **** **-**** ****.

Integrator ******

*** ******** ***** ***** ** * way *** *********** ** ************* ********** from ***-**** ****** *********, ***** **** providing * ****** *** ******. ******* from *******/*********** ** **** ********, ** the ****** **** ******* **** ****** of ******* ***********. ** ********, ****** have * ******* "**** ******", ** least *** ***, **** ********* **** Amazon ******* ***** ******* ******** *** drone, *********** ***** * ***** ***** guard **** **** ***** ***** ***** certainly ***** ***.

2017 ****** ****

*** ******* ******* ** **** ***** drones ***** *** ********* ** ***** 2017, *** *** *** ********* ** a ******** ****** **** ***.

FAA ************

******** **** ***** ******* **** ** "100% *** *********" **** **** ******, and **** **** ****** *** ********* registration *********, ** **** ** ********* 24/7 **** ****** ********** ** ****** drones ** ******** ******** ** ****** with *** ************. ********* **** *** be ******** ** **** *** ***** guards **-**** ** **** ** *** desire.

Compared ** ***** ******

****** *** *** ******** ** ***** terrain, ****** ** ***** ********* **** can ******** * ***** ***** ** * limited ****. *** ****** **** **** is **** ******* ******, *** ****** more ******* ***** ********* ******, ***** impacts ******-**** **** **** ****** ********** batteries ***** ****** * ****** *****.

***** ******, **** *** ******** ****, are **** ****, ******** ** **** not ********* *** ******** **** *****-****, having *********** *** ******** ** ** based ** *****-**, ***** ** ** way **** ***** ** ********** ********** a ****** ***** **** * ******** mall, ****** ********* ***** *** **** the ***** ** * ***** ***** of ***** ***** ******** ** *** times.

***** * ***** *** *** ******** and ******* ** **** ****** ******** along *** ******, *** ***** ***** need ** ****** ***** ******, ** land ** ** ********** ***** **** lets ** **** ******* ******** *** view ** *** ******(*).

Challenges *** ***** ******

******* ******* ************ ** ** **** task, *** *** **** ** * catastrophic ******* ********* **** *******, ******* this **** *** ***** ** * heavy ***** ******* ** * ****** and ** ******* ******** **** ****** will ** ****** *** *** ** any ***** ***** *** ******* ****** has ******, ****** **** ****** **** smaller *** *******.

*** "********* ******" ** ***** ************, as ** ** **** ******. ***** that ** ** **** ****** ** observation ******, ********* $**,*** *** **** in ********* * ****** ****** ***** ultimately ***** * ****** *** ** funding **** **** **** ******* ****** in **** *****.

Drones, ******, ** *******?

Comments (37)
U
Undisclosed #1
Oct 27, 2016
IPVMU Certified

Aptonomy says their service with be "100% FAA compliant" when they launch...

How so?

(1)
Avatar
Brian Karas
Oct 27, 2016
IPVM

I brought this specific question up with them in our conversation, as I was aware of this restriction also. They say they are aware of FAA regs, and how to satisfy applicable requirements. I do not know what the FAA considers acceptable, or when/where/how they may make exceptions.

U
Undisclosed #1
Oct 27, 2016
IPVMU Certified

You can apply for a waiver, did they say they actually have any in hand?

They're gonna need quite a few of them :)

(1)
(1)
Avatar
Mihail Pivtoraiko
Oct 27, 2016

Yes, the FAA allows this sort of UAS operations via waivers. A number of UAS operators have received waivers for night-time operations, as well as beyond line of sight operation.

Lots of other recent progress on this front, e.g. https://www.nasa.gov/feature/ames/nasa-plans-first-beyond-visual-line-of-sight-drone-demonstration-in-nevada

(3)
U
Undisclosed #1
Oct 27, 2016
IPVMU Certified

Agreed, you can operate if you have waivers, but you need to get them.

So, then the question is "do they have any waivers?"

(1)
Avatar
Mihail Pivtoraiko
Oct 27, 2016

No. To explain, the deployments of our alpha version of the product have all been done in compliance with standard regulations. However, we are actively pursuing the waivers for our first batch of beta users.

(2)
U
Undisclosed #1
Oct 27, 2016
IPVMU Certified

Thanks, Mihail.

Another issue would be the no multiple drone operator restriction, as it would seemingly kill the financial model.

Is there real hope that you can get a waiver for that?

Amazon, et al, have been trying for years now.

Avatar
Mihail Pivtoraiko
Oct 27, 2016

You are right, this is an important consideration. Since not all drones would be flying at all times, there is still a possibility to scale, but hiring extra remote drone operators would certainly be expensive. As soon as this challenge is overcome, we would be happy to pass the savings to our customers.

Avatar
Ari Erenthal
Oct 27, 2016
Chesapeake & Midlantic

I just talked to our drone specialist here, and he says all the specs here are perfectly plausible with existing, off the shelf technology (other than the on-board analytics that will identify a subject, which he couldn't comment on). The only limitations he could think of is the range, which is going to be a max of 4.3 miles from the base station, and the fact that base stations can interfere with each other if you're going to be using more than one at a time.

He also says that the that drone operator groups and drone manufacturers have been lobbying the FCC to allow autonomous flights, and that the drone community fully expects the FCC to issue new guidelines on non VLOS flight quite soon. It wasn't very long ago that the FCC apparently took the position that all drone flights were illegal unless the operator was licensed, after all.

(1)
(4)
Avatar
Mihail Pivtoraiko
Oct 27, 2016

Hi Ari, good point, off-the-shelf drones indeed have similar flight capabilities.

In contrast with hobby models that typically come with joystick radio controllers, the Aptonomy system would be similar to a PTZ camera (except that of course the camera moves in full 3D). A remote user would simply specify a location (e.g. by tapping on an overhead map of the property), and seconds later the visual data of that location would start streaming in. Upon arrival, the UAS orbits the location to provide 360-deg view; one can pause that and look around at will.

(1)
Avatar
Ari Erenthal
Oct 27, 2016
Chesapeake & Midlantic

Mihail, thanks.

How about the base station issue? My drone specialist is worried about interference if the user attempts to use multiple drones with multiple bases. Is it possible to have full coverage with multiple drones, or is a user realistically going to be limited to using a single drone at a time?

Also, he would like to know what the weight of the drone is fully loaded, if you can disclose that. He's assuming you're using the 15000mA LiPo battery?

Avatar
Mihail Pivtoraiko
Oct 27, 2016

Ari, the Aptonomy UAS are not subject to interference issues because they don't use standard RC joystick controllers. The user would connect to them via WiFi, LTE or a similar dedicated and secured data link. Much like two laptops next to each other can still use WiFi, so can Aptonomy UAS.

I'll double check with the engineers, but the DJI S1000 airframe we are using is rated for full takeoff weight of a bit over 20lbs. Indeed, 15000mAh all the way to 22000mAh are frequently used for this frame. DJI markets this airframe for high-end drone videography for shooting movies, commercials, etc.

(1)
Avatar
Ari Erenthal
Oct 27, 2016
Chesapeake & Midlantic

But then your range is going to suck. Isn't it? The wifi range on the S1000 is something like 30 meters. Even the Mavic's WiFi range is only 80 meters, and that's the longest WiFi range of anything on the market.

Avatar
Mihail Pivtoraiko
Oct 27, 2016

Ari, that's right, WiFi has limited range. If there is LTE coverage, we can use that. Otherwise, or if the customer prefers their own network for privacy reasons, we would setup a set of WiFi access points, an LTE microcell or a dedicated link, such as www.mobilicom.com. The system is agnostic to the data pipeline, but specialty solutions would probably run a little extra.

(1)
Avatar
Ari Erenthal
Oct 27, 2016
Chesapeake & Midlantic

Hmmm. If I'm putting up hotspots all over my property, I may as well put cameras near them at the same time. I'm not sure yet why Aptonomy drones are better?

Avatar
Mihail Pivtoraiko
Oct 27, 2016

Ari, good point. Just to reiterate, we have had a number of successful installations using 4G LTE (standard cellular network).

I think a surveillance drone would complement, rather than replace fixed cameras: a traditional camera sees one location all the time, while a drone covers a wide area periodically.

For example, if a fixed camera sees something suspicious far away, a drone can be summoned (even automatically) to go and check it out. Certainly, most analytics solutions do better with closer range and better picture quality.

In addition to surveillance, we have noticed in our deployments so far that the physical presence of a large drone has an important psychological effect that seems to really help with prevention. The present article correctly points out, however, that the current version of the system is better suited for sparsely populated areas, or in busy locations outside regular business hours.

I highly value your thoughts on our proposed use cases and the comparison/contrast with the existing solutions. Your feedback is very important to us.

Avatar
Ari Erenthal
Oct 27, 2016
Chesapeake & Midlantic

In addition to surveillance, we have noticed in our deployments so far that the physical presence of a large drone has an important psychological effect that seems to really help with prevention.

That's a great point, and one I hadn't considered.

UM
Undisclosed Manufacturer #2
Oct 28, 2016

Hi Mihail, very cool what your doing! Few questions:

Is your goal to stream video over LTE?

Thanks

Avatar
Mihail Pivtoraiko
Oct 28, 2016

Great, thank you for your interest.

Our current model supports streaming video over LTE and WiFi. Future upgrades will also support dedicated links, such as Mobilicom.

UM
Undisclosed Manufacturer #2
Oct 30, 2016

Knightscope tried to do the same thing and it didn't go well. They initially deployed the robot in a robust Miraki environment sharing wifi but then the customer kicked them out due to bandwidth. Then they went with 4G and had a $10,000+ bill.

Like Ari said, asking the client to deploy a wifi system for your drone and making sure you have perfect coverage will be a major challenge. LTE is also not cost effective at this point.

I know your products is futurestic, and these problems will be solved by time, but I believe these major challanges will post an issue going main stream.

I believe customers that will buy your product at this point are customers that struggle to have a person at this point (dangerous, remote, sporadic) not because you can replace a guard at this point.

(2)
Avatar
Mihail Pivtoraiko
Oct 31, 2016

That's a good point. The Aptonomy UAS is different in that all the AI is right on-board the drone, and it does not require much bandwidth beyond video. By far the biggest data pipe is for streaming the video (which could also be subsampled for real-time streaming, but available in full-resolution when the UAS comes back to base).

An interesting post by Brian on a related topic a couple of days ago: https://ipvm.com/reports/digital-barriers-tvi

U
Undisclosed #1
Oct 31, 2016
IPVMU Certified

The Aptonomy UAS is different in that all the AI is right on-board the drone, and it does not require much bandwidth beyond video. By far the biggest data pipe is for streaming the video...

How is the Apotonmy UAS different than Knightscope?

Knightscope has AI "right on-board" as well. Regardless, the "biggest data pipe" in both is certainly for streaming the video, no?

Avatar
Mihail Pivtoraiko
Oct 31, 2016

Absolutely, in terms of networking and integration, our UAS looks just like a standard wireless surveillance camera (except of course it moves in full 3D).

(1)
Avatar
Craig Kodish
Oct 27, 2016

Very insightful Ari. For the "drone doubters" out there citing current regulation as a limiting factor for future success, I would consider how quickly deregulation can occur and then shock industries. Recall the Airline Deregulation Act of 1978.

Although $120K in seed funding probably won't finance an effort to yield regulation changes in Washington, I sure wouldn't mind the big boys like Amazon, Google, and GoPro spending their lobbying dollars to promote my agenda.

(1)
(1)
U
Undisclosed #1
Oct 27, 2016
IPVMU Certified

Hi, Craig!

Believe me, I'm no drone doubter, and I own several...

I think regulations will change, and quickly.

However, the article says

The company expects to have their drones ready for customers in early 2017...

and that

100% FAA compliant when they launch

Which implies that you think that new regs that allow operators to not have VLOS are just months away.

And the latest regs just came out in Aug.

Moreover, seeing that the number of changes necessary that will be required to run your service as outlined above is numerous, don't you think that the service will likely be modified in some manner by the time it gets to market?

So, why quote prices on something that is not legal today and may never be offered in the form you propose?

(1)
Avatar
Ari Erenthal
Oct 27, 2016
Chesapeake & Midlantic

Yeah, they seem real optimistic.

Avatar
Mihail Pivtoraiko
Oct 27, 2016

Guys, great comments on compliance. Trust me, we think about that all the time.

I am really thankful that the FAA recognizes the importance of drones for our economy and has made sweeping changes in August. I understand that further improvements are to come soon. In the mean time, the FAA waivers is the procedure we rigorously follow to stay compliant.

(1)
Avatar
Craig Kodish
Oct 27, 2016

Pleased to meet you Undisclosed #1. Just to clarify, I am not am employee of Aptonomy, just an admirer of their entrepreneurial spirit and believer that their product will fill need in a developing market.

I have no new information to estimate when deregulation could occur, like most people on this thread, I have observed an increase frequency in discussions about drones, security, and regulation.

To answer your interesting question about why an organization would assign a price something that is not legal today and may never be offered in its current form, my thought is that an entrepreneur must develop a Minimum Viable Product (MVP), a product with the absolute minimum number of features that customers will actually pay for. It is the responsibility of the entrepreneur to make a winning pitch to Angel Investors or VC's like Y Combinator for seed funding. VC's want to know that their investment will yield above normal returns, so the entrepreneur must share a business model which highlights revenue and cash flow estimates. Without a price and estimated number of sales, there is no formula to estimate revenues and expected return on investment.

(1)
(1)
U
Undisclosed #1
Oct 27, 2016
IPVMU Certified

VC's want to know that their investment will yield above normal returns, so the entrepreneur must share a business model which highlights revenue and cash flow estimates.

Yes, agreed, the price is for the VC's.

So render unto the VC's what is the VC's. And skip the firm public pricing of the gelatinous offering until things congeal.

(1)
Avatar
Mihail Pivtoraiko
Oct 28, 2016

The price is for everyone -- this is simply the rate at which we plan to deploy the system in early 2017 (as the article mentioned).

U
Undisclosed #1
Oct 28, 2016
IPVMU Certified

With so many unknowns about exactly what service you can provide (non-VLOS/Night-time/flying above people) and uncertanties about what that will cost you (1 pilot/many drones vs 1 to 1), I would think its more of a target price than what the actual price will be.

But, if you're saying its a firm price, I'll accept that.

Avatar
Mike Dotson
Oct 27, 2016
Formerly of Seneca • IPVMU Certified

One gotcha in this scenario is when the 'bad guy' has a gun.

This thing is slow enough to offer a nice target and YouTube has plenty of videos of drones getting shot out of the sky.

The operator will need to keep the drone moving quickly like they do in the 'drone races' to help mitigate this issue.

The 'robots' that are also discussed in this thread also have the same sort of problem.

(1)
Avatar
Mihail Pivtoraiko
Oct 27, 2016

Mike, good point, fully agree with you. While of course it's infinitely better they shoot at a drone, rather than an on-site guard, the legal consequences would probably be lighter, should they get caught.

This and other preventative measures are designed to stop the bulk of the attacks. However, responding to a heavily armed, sophisticated intruder would require additional resources, including law enforcement. An Aptonomy UAS could be a valuable tool to provide situational awareness to the authorities, perhaps at the risk of losing the vehicle.

(3)
Avatar
Meghan Uhl
Apr 17, 2017

One gotcha in this scenario is when the 'bad guy' has a gun.

This thing is slow enough to offer a nice target and YouTube has plenty of videos of drones getting shot out of the sky.

 

This particular "gotcha" is not just a risk to the UAV but also to live guards &/or traditional camera systems on a site.  Bottom line is that you're not going to achieve 100% protection to any site against theft, vandalism or death by bullet.  The goal is to create an environment of protection that will deter &/or capture the bulk of the bad guys. For those that bring a gun & commit "murder" of the drone, guard or camera, at least the last thing the drone or traditional camera did before its demise was capture the bad guys image and send it to the cloud or monitoring station.  A dead live guard takes that image to the grave with them so bringing the shooter to justice is much more difficult for PD.  Deploying video technology of any kind is a compliment to live guards and provides more safety & security for the guards as well as the customers assets.

(1)
U
Undisclosed #3
Apr 17, 2017

Perhaps DJI should attached a cable to feed power constantly to the drone along with a 12 pound load. Then let the drone run continuously in various hi and low temperatures/weather so the over all lifespan can be calculated.

Perhaps you already have a mobile base charging station (flat bed truck) so you can deploy anywhere.

Will DJI develop a butane/battery hybrid drone? (est. 3 hour flight time).

Is there a tow drone? Configured to go out and retrieve a fallen drone?

What if a drone was equipped with a fuselage of paint and programmed to "heat sink" all active cameras with IR? It can fly around a facility and paint the cameras FOV. Perhaps the use of IR on the drone can give it the ability seek and destroy cameras that do not have IR. You can send in 3-5 drones to a facility and knock out 20-30 perimeter cameras. (always a bad guy, I know).

Perhaps one could make "The Owl" a drone that can intercept and disable any of the DJI type drones in flight. Perhaps jam and rebroadcast the GPS signal so the drone is attracted to the Owl. Then drop a nylon spider web onto that said drone.

Perhaps just thinking about drones is fun! Good Luck Aptonomy I am sure within a decade drones will reign...they already do.

JH
John Honovich
Apr 17, 2017
IPVM

Perhaps DJI should attached a cable to feed power constantly to the drone along with a 12 pound load. Then let the drone run continuously in various hi and low temperatures/weather so the over all lifespan can be calculated.

Related, there is a company developing / marketing drones for surveillance that uses tethers - CyPhy. We have a post on them in the queue.

(1)
Avatar
Meghan Uhl
Apr 17, 2017

Problem with CyPhy is that they are like $200k so out of reach for most security applications.  If Aptonomy can achieve all its stated goals i.e. FAA waivers etc., they are much more in line with the security market from a pricing standpoint.  I believe they are at approx $20-$30k per drone.