As often happens while engaged in a mundane task, my mind wandered while I was mowing my small suburban plot of green this weekend. “Why, in 2017, am I still mowing the lawn?” In a lot of ways we’re living in the future — we walk around with fantastically powerful computers in our pockets, some of us have semi-autonomous cars, and almost anything can be purchased at the touch of a finger and delivered the next day or sooner. We even have robots that can vacuum the floor, so why not a robot lawnmower?
It turns out we do have robotic lawnmowers, but unfortunately, they kind of suck:
Bearing in mind that the video was produced by Husqvarna, it should come as no surprise that their entry in the robotic lawn care field was the top performer, and that other variables that would likely have challenged all the mowers equally, like tall, wet grass, were not tested at all. What I saw in the video was a bunch of mowers that all suffer from a couple of basic problems that answer my question of why we don’t see robotic lawn mowers in every suburban yard.
More Power!
You might think my main beef is the universal need for wires to define borders for the robots. It would be nice to see a fully autonomous robot that could locate the flower beds just using GPS and image analysis, but wire boundaries seems like a reasonable means to the precision needed to keep the bots contained. Wire boundaries are akin to “invisible fencing” for dogs, and the one-time cost of installing the wire would be amortized over years of having a robot handle one of my weekly chores. That’s a different article — [Will Sweatman] already did a great job covering automonous robot location tracking tech.
The real problem I see is a lack of power. Each of these robots is powered by batteries and they seem terribly underpowered both in terms of drive train and cutting. The lack of power is most evident in the inability of most of the tested units to deal with hills, but some were not even especially good at cutting grass. A lot of the cutting ability seems to have to do with the compromises made in the blades’ design — there are no massive steel blades on these bots. The most successful are essentially whirling razor blades that won’t survive a real-world yard where rocks, tennis balls, fallen tree branches, and pine cones the size of a chihuahua are common obstacles.
There’s a reason your average walk-behind lawn mower uses a fire-breathing internal combustion engine: nothing beats it for concentrated, portable power. The blade spinning beneath the deck of the mower has an incredible amount of energy behind it, enough to cut the grass, mulch it into tiny bits or blow it into a bag, and still have enough left over to move the machine around so you don’t have to. As noisy and environmentally unfriendly as it may be, the internal combustion engine is king of the greens, and even though battery and motor technology has come a long way, it’s hard to see that they’ll be able to match a gas engine’s power to weight ratio anytime soon.
Safety
To me, then, the essential question is: how do you leverage the concentrated power of internal combustion for safe, effective, automatic lawn care? I doubt we’ll ever see consumer-grade lawn bots with gas engines, primarily because automatic refueling is so much more complicated with liquid fuel than with electricity. The inconvenience of needing manual fueling coupled with the costs of making the machine tough enough to stand up to regular use in a dynamic environment with vibration, heat, oil and fuel spills, dust, and moisture pretty much make a consumer-grade gasoline lawn bot a non-starter.
There would be a ton of problems to solve before “Lawn Care as a Service” ever becomes a reality, not least of which is assuring homeowners that a fleet of powerful robots swarming through their neighborhoods with spinning blades of death is a good idea. But there seem to be powerful economic forces at work that could prompt a sufficiently forward-looking lawn equipment manufacturer to start working on a “big boy” machine for the professional market rather than turning out any more puny lawn-Roombas that are destined to fail.
I’ve wondered about combining a robotic mower company and a dog “invisible fence” company so you’d only have to lay one wire.
That being said, as a long time pet owner, I do NOT recommend invisible fences. Dogs can break through them if they’re sufficiently determined, and they don’t keep other animals and children out. They’re basically useful if you’re outside with your dog and keep tabs on them, not for sticking them in the yard and forgetting about them.
That already exists. Many of the roomba style lawnmowers utilize a buried wire that they can sense. I forget if it’s energized or some sort of permanent magnet arrangement.This keeps them out of flower beds and vegetable gardens with varying degrees of success and increased cost to the owner dependent on the perimeter length.
I think you misunderstand me (or I said it poorly). The mower company needs also make a dog collar that’s compatible with their buried wire so you only need one wire for your dog and mower.
Yes, definitely! And then the robot mower should be able to retrieve the dog, should it break out (again). That sounds like fun!
And de-claw the cat as well just for good measure!
That’s precisely the way I’d do it, use the invisible fence wire. I like the article’s suggestion of LCaaS though, something to consider.
Also I’d make it as a kit that is compatible with existing self-driving mowers and garden tractors. A small battery-powered unit that operates the brakes, gas, and wheel to pilot your existing mower through the lawn, using your dog’s invisible fencing wire as a guide. Can be removed from your tractor/mower with a few clips. It’d have the same safety concerns as the current generation of robot mowers but also a large physical panic button and phone app with a panic button.
Oh and by the way, gas-powered remote control (non-autonomous) mowers are a reality today. See below. But as of yet no one makes a kit that straps onto your existing John Deere to make it mow automatically.
http://www.slope-mower.com/
Better yet put a collar on a sheep, cow, or goat and let it keep the grass short.
I understood this, but I thought: Are you sure, that that the perimeter fence for the mower and the dog can be at the same location? I think there will be places the dog should not go, but the mower and the other way round.
Gasoline is hardly the only option. You can buy a self propelled cordless mower .today that swings a big steel blade. Worst case – an autonomous mower doesn’t have to mow the entire yard all at once – it can nip back for more juice and then pick up where it left off. The real issue for commercial models is price. The cheapest Husqvarna is over $2000 – has to be a huge mark up.
Huge markup or huge liability cost.
You don’t really need to swing a big steel blade. A garden strimmer(UK), weed-wacker (USA) or whipper-snipper (Australia) will do a fine job of cutting grass with a 2mm plastic cord.
Sling-blade?
I call it a Kaiser Blade. Mmm hmm.
No, it’s a Sling Blad.
Exactly. Steel blades are used on gas mowers because they act as a flywheel, and the gas engines turn much slower than an electric motor would.
Actually tip speed is typically limited to around 18000 feet per minute or ~200mph for safety reasons. This can be around 3000 rpm depending on the blade size of course which is quite easy for an IC Engine to reach.
As great as this idea sounds, I will believe it when I see it. There’s a huge number of technical, legal and practical hurdles to overcome. Self driving cars at least can see the road and are already quite expensive compared to a lawnmower. How does it differentiate from items trapped in the grass like a turtle or toy? How does it track it’s location? Or know where it has been? How does it deal with avoiding traffic? There’s no legal framework. It’s a huge liability even if it was low enough cost.
Can’t we genetically engineer grass to only grow so long instead? Pubic hair only grows so long, so it certainly feels possible.
The last would be interesting. The benefit doing a lawn vs driving the road, is a lawn is relatively static. Map the entire thing, and place precise markers here and there to keep things on track so to speak. The rest is dealing with stuff that’s not suppose to be there.
Much like a Roomba, the lawn mower doesn’t need to see where it is going, only to know that it is within the area it’s supposed to be… then go straight until you hit the edge of that area, turn around and go the other way.
Much like a roomba, if you don’t clean up the area before it runs things like childrens socks will gets sucked up and break things.
Much like a roomba, if it doesn’t get a particular location today, it will get it tomorrow… it makes up for it’s inefficiency in it’s pattern by running regularly instead.
As for genetically engineering grass, there is no mow grass available… but it doesn’t stop growing so much as just flops over at a certain height…
Clover is a decent option too if you want green.
I pulled out a good section of my yard (500 sq ft) and replaced it with river rock… still need to “fertilize” once a year but now it’s with weed preventer…
Have you considered using moss in your rock garden?
I don’t think I would want a pubic hair lawn. Wouldn’t it be susceptible to “crab”grass?????
I’ve been questioning the whole concept of a lawn. If we plant clover the lawn would be self fertilizing. Plant cultivars of Buffalo grass (e.g., Cody) and you limit growth height and the lawn is drought tolerant. Why do we traditionally pick such a resource intensive monoculture lawn?
It feels good, looks good, and you can rub your bare feet in it to get vitamins.
Neighborhood pets fertilize it for you.
Well, the original purpose of having a lawn was an ostentatious display of wealth: “Look, I have so much land, I can afford to cover some of it with a useless plant that provides no value, only “beauty”, and takes resources to maintain!”
Some of us haven’t moved on.
The trouble with using a ground cover that only grows a certain height is getting it to also grow thick enough that it will choke out weeds without regular mowing. I’ve seen a moss lawn transform into a lettuce and onions lawn from being allowed to sit neglected too long. No, seriously, we get some rather weird weeds around here.
That is why God made 2-4-D.
“But what about a pro-grade autonomous machine? Landscapers already pay a lot of money for big, powerful machines that cut grass quickly and efficiently.”
SCAG here, but we have a LOT of lawn so it’s worth it. And yes I’ve entertained the idea of an autonomous mower*, but the grit of it is a person is a lot cheaper and more flexible in dealing with our terrain. Even multiple people.
*I imagine the geek in most would borrow the technology from autonomous farm equipment and scale down accordingly.
I think the paradigm could change if such a mower could maybe take longer to cut the grass. Instead of trying to cut the whole yard in less than an hour, why not have a device that takes a few hours or even days to cut the yard? Assuming you could keep people from stealing it, just leave it outside 24/7 and it would slowly wheel around and very slowly cut the grass. Now the blade doesn’t need to be a huge swirling blade – it could potentially just be a string trimmer or a hedge clipper or something that enables a longer battery life. It can have multiple recharge cycles before it finishes the lawn, if needed.
That’s what they already do.
The Huskavarna, Bosch, Honda, etc just spin tiny razor blades effectively trimming small bits off the tips of the grass over the span off days. They are very light and quiet. Most have a base station they return to automatically when they need charged. Once the charge is complete they return to the mowing if it’s a scheduled mow time and not raining. I have mine scheduled to come out at night. This way I don’t have to worry about someone stealing it or kids trying to pick it up. Not that I would have to anyway. These things stop the blade faster than I can pick it up and get my hands under (I’ve tried). The anti-theft mechanisms are meh. No one cares about loud noises. Best thing is just scheduling it to come out at night.
Since it randomly goes around the lawn and never takes very much off the top, your lawn is typically healthier. Just factor in the cost and frequency of blade replacements. Some models require you to replace their proprietary blades once a week!
You’ve just described a goat.
Seriously, in Norway they use goats in some cities to keep the green areas under control.
Ideally grass would be genetically engineered to not need mowing.
Short of that.. Robotic mowers are fine.
Problem is, people like me.
I got the cheaper version because, given what they are, they’re too expensive.
Then I found out..
..you NEED a rain sensor
..you NEED it to handle a 30degree slope
..you NEED it to have non-proprietary blade replacements in case the company folds (my first two did)
..you NEED it to recharge itself
..you NEED it to handle a perimeter wire signal loss due to power outage
..you NEED it to operate off a schedule
..you MIGHT NEED it to have GPS
etc, etc, etc…
I thought all of those were frills. Nope, they’re the difference between a tool that replaces your mower and a hobby project that costs time/money.
Although those features are only available on the >$2k mowers, consider what your time is worth and >30hours a year wasted doing it manually.
I just got my Husky this season (my fourth robo mower) for ~$2k. If it lasts me this season, I’m spending $67 per hour! If it lasts five, I’m spending $13 an hour.
Considering the excessive cost of what amounts to an autonomous RC car and the amount of wasted labor spent on lawns, I’m surprised no one has done an open-source version.
Super 3D printable. The brains could be as simple as a 8bit PIC.
No $67 an hour you might as well do it yourself or pay the kid down the street to do it.
Or someone can have Putin drop a nuke where your at the world would be a better happier place and I don’t think anyone would send a response.
They have its called ardumower Google it
ardumower?
Sorry. I should have been more clear. I meant a robotic mower that satisfies all the issues I hinted at earlier and costed less.
Considering the intent of a robotic mower is to save time, spending time to build something that has no warranty, is more expensive and less capable sounds silly.
I guess it’s a good starting point though. Cost could be reduced and capabilities improved.
If you’re looking for autonomous lawn trimming, why not rentagoat.com ?
The free fertilizer is not a wanted feature.
Lawn care as a service has been around forever. I pay a company ~$60/mo to mow my lawn every other week.
This is the comment I came here to see… Poor choice of wording on Dan’s part I’d say.
Either Dan chose his words poorly or commenters didn’t read the article; I’m betting the latter because I sure got the idea. He’s talking about upgrading the existing lawn care as a service that has been around forever, making it more efficient.
Meh! I’m with [biosehnsucht] and [Allen]. Lawncare is already a service. You can’t buy “lawncare” — it’s not a good. It’s a service. You non-economists get off my lawn!
But yeah. The point is that a good application of expensive, autonomous mowers would be helping the already existing lawncare sector. I buy it.
Here in Poland we just hire some ukrainians. For 5 zlotys an hour ($1.30) they’ll gladly cut grass with bog standard gasoline mower.
Here in Sweden we have poles.
Here in the South we have slaves.
You need a line of products and the yard will look like a miniature wheat farm during harvest. An autonomous clippings collector that takes the cut grass to an autonomous conveyor belt to your “silo” or compost or yard waste bin.
Or a mover that just cuts first them docks and picks up a sweeper attachment to collect clippings.
Personally, I think minions would be a lot more fun.
I’d love it if front lawn food gardens were more popular, or even permissible in certain HOAs and cities. Food, not lawns.
I’ve had a Robomow RL-500 since 2001. The machine had 300 hours on it the last time I checked. It needs a new set of batteries every few years, other than that it has been trouble-free.
only 300? when i was working on the husky automowers highest score was (no joking here) over 9000 hours!
I have a 2nd generation husquarna since ~5 years and so far the lawn has always been in perfect shape and maintainance is like 6 hours a year (replacing the blades once a season and permanently having to pick up the small branches from trees that fell down. Saves me 2 hours a week.
you want to have it switched off when sleeping on the lawn though, the blades are razor sharp. So far “dolly” killed a purple spraycan, a bottle of chlorine bleach and countless beercans and underwear.
That must’ve been a hell of a party!
Lol
Why have a lawn? They use some much water during the summer, and during the winter they do nothing.
Agreed. Nothing says “conspicuous consumption” quite like a private lawn, maintained at great expense and serving no purpose other than giving you something to tell people to keep off of.
This conversation was beaten to death a few articles ago. For some people it’s not an option to forgo a lawn. I will be leaving my HOA neighborhood soon as I can and planting crops at our new place.
A friend of mine calls a lawn “a crop you have to harvest once a week and then throw away”.
I harvest mine once a week, compost the clippings, mulch the garden with the compost, and eat the vegetables. There’s a lag time, but I end up eating my lawn. Just saying.
Important aspect of robotic mowers for this site would be ‘hackability’. Seems that the husqvarna automower has a large hacking community and a serial port. Thats all we need folks!
where? i want a secondhand one, but the dealer only software is a right pain.
Militaty drone aircraft illustrate that the technology is available now to have remote operated gas powered commercial lawn mowers with “pilots” located in some other country with low wages and no overtime rules. Seems like a cheaper solution to developing autonomous lawnmowers.
Then again, I figure that is the same solution that will win out over autonomous cars: remote chauffeurs.
I would not set my live on a remote data connection. I also don’t use cloud services if avoidable. The chauffeur has to be on board, be it silicon or wetware based.
I remember reading about something like that and digital immigrants seems a lot more likely than strong AI becoming common place as truly smart robots has turned out to be a difficult problem to crack.
But remotely controlling a dumb machine from halfway around the world is just off the shelf technology.
https://www.newscientist.com/article/mg22429982-300-will-remote-controlled-robots-clean-you-out-of-a-job/
I had LCAAS — it came in the form of the teenager that lived next door. It was good for 5-6 years; then he discovered girls, and that was the end of that.
LCAAS ver 2.0 (beta) lives in my house, and is almost old enough to start the mower on his own without maiming himself.
I already have this, I pay $40 a month and the lawn is magically cut every 2 weeks. I assume someone comes by with a laser and trims it. There is no reason in 2017 to cut your own lawn.
I just let the sheep out back for a few days and they eat most of the stuff down. The little bits they don’t like I can weed wack. Mowing is such a waste of resources. Another really bad idea that spread fast.
Self plug here, with MowBotix. We are working on the exact problem for lawn care and golf industries on electric/gas/diesel mowers. For those that are wondering about safety we are taking that seriously. Here is another video of us starting our safety system using a combination of point clouds, 2D LIDAR, and sonar. Here is a video showing only the 2D LIDAR.
https://youtu.be/Be8IzfQfuO8?t=1m22s
More videos here https://www.mowbotix.com/media/
Impressive! Instead of auto-refueling by refilling the unit’s fuel tank, may I suggest swapping the primary fuel tank itself for another, full one. A smaller, secondary tank would hold enough fuel to make sure that the mower continues running long enough to swap. A number of full (and empty) ones can sit on a rack, and the operator can just go by whenever needed to refill the empties.
I dream of electric sheep. Screw mowing, watering, weeding. Synthetic lawns is where its at!
The question is, do androids dream of electric sheep?
My lawn is small, I can’t get past the idea that a raspberry Pi with OpenCV looking out of the upstairs window could easily direct a pretty dumb robotic mower around the big green bit whilst leaving the not-green parts unmolested.
For a big lawn, perhaps a Pi up a pole or few cheap wifi webcams could achieve the same result.
To be honest this sounds more like a solution looking for a problem as lawn care as a service is as old as lawns.
Though I just bought a ZTR mower so it’s just a 20 minute job for me.
How about ask [Colin Furze] to trim the high grass with his hoverbike:
http://hackaday.com/2016/04/28/colin-furze-flies-the-dangerous-skies/
For navigation perhaps one could you use a small lidar unit and a series of bright reflectors around the yard (enough to triangulate on). This technique is used to stitch together scans from high range terrestrial scanning lidar units. May be some deep learning and a training mow with the unit being pushed by hand.