Sunday, May 17, 2009

iRobot Ember Ushers in Era of Military Microbots

Updated with video below!

EXCLUSIVE, BREAKING: iRobot Corp. (NASDAQ:IRBT) has developed a new microbot for military applications -- the paperback-sized iRobot Ember!

This hot new bot, still in the prototype stage, is featured on iRobot's new Facebook page for its PackBot family. As you can tell from the photos iRobot posted on the page, Ember resembles a miniaturized PackBot, complete with iRobot's flipper technology allowing it to right itself and climb over obstacles.
The iRobot Ember was developed under DARPA's LANdroids program, which is intended to create teams of tiny robots for military applications, including, as the name implies, setting up an ad-hoc network of hotspots. The robots are intended to be about 1 pound each, be smart enough to detect and navigate around obstacles and ultimately cheap enough to be considered disposable.
There are, of course, numerous movie precedents for such robots (remember the message carriers in Wonder Woman or the tiny 'bots on the Death Star in Star Wars). One can certainly imagine a time in the not-too-distant future where every soldier will be handed an M-16 and a suitcase with a 3-Pack of Embers (see photo below!). If iRobot can get the cost down to the military's target of $100 -- which I think is a HUGE stretch given that the PackBot is 1,000 times more expensive -- that'd be a hell of a bargain. I know I'd want a few! Other tidbits that could lead to add-on technologies -- the 'bots include USB and SDIO interfaces, so, just like big brother PackBot, this little guys could conceivably have attachments like webcams, microphones, bomb sniffers, etc. plugged in and ready to go. How soon until this tech gets previewed on a show like "24"?

We wrote about iRobot's $2.5 million, three-year contract award to develop the LANdroid last year. Note that several other companies also got awards, so if this becomes a shipping product, iRobot will have competition. But my guess is that iRobot's flipper technology and sophisticated software should give them an edge.

According to iRobot:

Ember is a prototype and was developed under Phase 1 of DARPA’s ongong LANdroids program.

LANdroids Program Goal:
To provide warfighters operating in dense urban environments with tools to deploy and maintain an ad-hoc communication infrastructure.

Ember’s Statistics & Capabilities:
--Approximately one pound
--Unprecedented mobility for vehicles of similar size
--Uses a flipper mechanism for self righting and obstacle climbing
--Accepts USB and SDIO based payload modules, including radios and sensors

One single warfighter will be able to carry and deploy multiple robots:
--Ultimately, these robots will be inexpensive to the point of being disposable
--They will be robust enough to allow the warfighter to drop and throw them into position
--They will be smart enough to detect and avoid obstacles while navigating its environment
Here's the three-pack pic I promised you:

Here's also a graphic of the ad hoc network these bots are supposed to provide:

UPDATE: Our friends at Wired have posted the following video of the iRobot Ember in action:

Tags: , , ,


Mark said...

Love the Photoshop diagram of the mesh network the wee robots will achieve with their tiny little antennae :-) Not sure if you'll get city coverage with them... more like line of sight up to a few hundred metres. Certainly not very far if the bot's antenna is nestled underneath a nice big metal railway line :-)

mlebauer said...

Very cool robot, but it's technology in search of a realistic application.

Battery life will be limiting. So will wireless range and computing power for the envisioned wireless LAN, as mentioned above (although the concept is that the bot would move away from the railway line to improve signal propagation, it wouldn't improve much with the antenna 2 inches off the ground).

I work in the space and can't imagine the wireless hardware costing less than $2-300 by itself, and that would be using commonly available 2.4GHz WiFi chipsets. Radio signals propagate best when elevated. The antenna on the back of Ember is not only low to the ground but would have low gain. In cities you would rarely have line of sight. To keep cost and power consumption low the radio couldn't produce much power, max 100mW (comparable to a small home wireless network). All told, range between Embers might be 100' at best. I can't see how this concept will ever fly.

However, if you eliminate the robot, solders could place low cost wireless nodes with a higher gain antenna on a small mast. Range would be vastly improved, power increased, and all battery power put towards higher power radios. I could envision such a system for less than $1000, approaching $500 with volume. $100 would still be a stretch.

Presto! Much less need to worry about the shading effects that Ember is supposed to address. That concept is achievable with off the shelf technology. DARPA should have saved their money.

thorn_stevens said...

mlebauer, I think you may have a point - except, who is going to install those antennas? The point of the iRobot Embers is they can go in first, not risking a soldier, and are meant to be so cheap, you could send them in by the dozens.

Even if Ember never is used as a WiFi node, lots of cool stuff is generated from DARPA research, and I can easily see Ember having many, many other uses that could be valuable to the military (as well as spy agencies).

mlebauer said...

Sure it's a cool device, maybe it could serve as a client to a wireless network. But the network itself?

For that they use a cheap telescoping mast with tripod and integrated antenna and attached radio. True, someone would have to set it, but that wouldn't take much time. Range would be so much better that the soldiers wouldn't have to get as close to the action. The battery could last far longer without the power hungry robot.

One article I read said that Ember's current cost is 100 times the target of $100, without radios, sensors, or other envisioned accessories, or $10000. DARPA's target is $100 for lots of 5000, including radio & sensors. How can they possibly get close to $100?

The main problem is that with minimal range 5000 would be needed to cover even a small town of 1 sq mile, given the expected range between Ember nodes as conceived of max 100 ft. The batteries wouldn't last long, needing to power both the robot and the radio. There's no way for cheap networking hardware to scale to that level. The concept won't work.

The concept I described will work albeit with somewhat higher risk to the warfighter, while the LANdroid concept won't. Drop the radios from an MRAP armored vehicle and the risk declines substantially.

bethesolution said...

The paramters established for this project have themselves created the barriers that will ultimately cause its failure. Lets examine. Parameters of $100 each, weighing 1pound, transmitting positional coordinates, micro antenna, and pocket sized.

First to transmit data over a reasonable distance requires a higher radiated output, which in turn requires a large amount of power. This power requirement means the batteries would take up the majority of the available room. Second, the batteries themselves would weigh greater than a pound themselves.

Third, the micro antenna also means a limited transmission range. Although higher frequencies require smaller antennas their sizes also are subject to signal antenuation through dense materials. This means in urban environments you would require a higher placement of receiving antennas.

The flipper design requires bursts of current to lift and flop the robot thereby reducing the batteries performance. Their depiction of potential coverage throughout a city is more fantasy than reality. Company's like irobot are constantly being funded for these unrealistic products because of their political influence rather than any real criteria. Smaller unfluencial companies are scrutinized while the large "good ole boys club" members are consistantly funded. If a smaller organization submitted for funding with these parameters they would be laughed out of the DOD.

What a waste of money!