Comment by msandford | Hacker News Reader

Comment by msandford | Hacker News Reader

Parent Full thread

How much does a pallet cost to manufacture and how much does a lo-jack cost to implant? Further, how much does it cost to keep the lo-jack running for a decade?

I used to run a 10,000 sqft logistics facility (really tiny in that world) and we kept a stock of about 20 pallets. I seem to remember the wooden ones costing $2-$5 used and the plastic ones less than $50 new. The most expensive pallets I could find from U-Line are $100 or so. http://www.uline.com/BL_8204/Rackable-Pallets

Figure that adding an armored place to put a lo-jack would cost $5-$10 (gotta protect it) and that the actual lo-jack would cost at least $50. It would need to be metal encased with a way to sneak an antenna out, have a "lifetime" battery (lithium-thionyl-chloride, not cheap http://www.all-battery.com/lithiumprimarybatteries.aspx) and the whole thing would have to be molded in to the pallet lest someone get clever and remove them cheaply. The batteries MUST be an integral part of the lo-jack so that someone couldn't just remove the batteries and neuter the device.

Then the data plan to keep it connected up to the world via the internet would cost at least a buck or two every month (in HUGE bulk purchases, individually it's something like $10/mo/device). Worse is that if someone puts them inside a metal building or metal shipping container or metal truck they're going to have a hell of a time getting a GPS fix and/or communicating with the cellular network.

So now that you've done all this your $40 to $100 (max) pallet now costs at least $50 and perhaps even $100 more PLUS it costs you something monthly whether people are renting it or not. What do you get to charge for this? From when I can tell, about $5 per trip. http://www.palletforum.com/tm.asp?m=26458 That's on a nominal $20 pallet cost although they may cost more now, the forum post is 2 years old.

I can't see how building a lo-jack into a pallet is going to be a compelling value proposition given all the limitations.

I work for an insurance company. The company currently pays for lo-jacks for high-end cars, and has often studied the cost of installing some variant of it for the entire fleet.

A GPS car tracker costs a lot less than what you're quoting, when purchased in bulk from China (I don't know about reliability figures though :) ). I think it was in the U$ 15 range for the entire thing. Certainly not the armored thing you're describing, which sounds pretty expensive.

And the cell phone data plan could be negotiated way down (I think your estimate is correct, about U$ 1 or 2 per month). What I pay here as a consumer in Uruguay is less than U$ 5/month for 384 megabytes. Getting a fix is definitely a problem, but you can track it until it gets inside wherever it lost the signal (hopefully it's a warehouse).

That said, there's a lot of low-hanging fruit to start putting GPS trackers on pallets :) (I still don't see them being ubiquitous with pets and other higher-value items to track).

If it wasn't feasible for cars (MUCH more expensive than pallets), then I agree with your estimate that it won't happen soon, but it's going to be possible sometime in the near future (and somebody is going to be ready for that, and make a killing).

Yeah $15 for the electronics. $20 to $100 for the battery depending on how good your hardware/software is. Probably a cast aluminum case, and then some way to get the antenna out. Plus the whole thing has to be rugged enough to survive the injection molding process which is no joke.

Worse is that even if the ENTIRE lo-jack is built into a single monolithic chip (which will take millions to engineer) you still have to power the thing. The transistors might be getting exponentially cheaper but the batteries aren't. And you need quite a bit of batteries to make this happen.

If you can get a fix to a warehouse you're OK but once the pallets go into an intermodal shipping container and stacked, you're SOL. Shipping container bottoms are made out of wood, not steel so sitting up on a truck they can still get signals. But once they're stacked you're in a world of hurt. Sure the containers might pop back onto the grid once they get unstacked but that might happen once they're out of the country, potentially to countries where you don't have subsidiaries who can go get them back.

So you have two fixed costs: batteries and data/cellular. You have to buy these and you have to pay for them on an ongoing basis. I think what would end up happening is that your battery and data (and plastic recycling and lo-jack refurbishing) costs will end up eating whatever profits you might manage to eke out such that it's not that great of a business to be in.

Aren't shipping containers steel all around? They do have wooden flooring though, but under that is steel from what I understand.

I'm not around shipping containers day-in and day-out but I don't believe they are. There's SOME metal down there to hold them together but as I understand it they're purposefully not 100% so that if something spills it doesn't pool in the bottom.

http://www.youtube.com/watch?v=KzWAj_gea7o

IMO, one random update per week is good enough to get most of the value from lojacking the things which would drastically lower your power needs. Probably not low enough for parasitic power draw but low enough to spend less than 5$ on the battery.

Just thinking out loud here--what about just using RFID for most of the pallets, and then putting powered electronics w/GPS only in some fraction of them? Say, 1 out of 20 or 50 (or whatever) pallets would have a battery, GPS, and an RFID reader; then those could read the RFIDS on any nearby RFID-equipped pallets. And transmit their location and the nearby RFIDs back to home base.

That's an interesting idea,and while RFID can be read up to nice distances(16 meter) but the RFID is quite power hungry and is dependent on antenna quality to achieve that , so it becomes a very challenging task and maybe an impossible one to build a decent concentrator.

lojack won't work for the reasons you mention, RFID really is the best solution because it's cheap and they don't need to change their existing inventory. There's no need for plastic pallets, they could embed the RFID tags into each of the wooden blocks which make up the corners of the pallet. If one doesn't already exist, it's probably quite simple to produce a nailgun like device which literally shoots an RFID tag into a block of wood, which would make tagging existing pallets straightforward and cheap.

The palettes are cheap enough that they actually don't need the exact location of each of them. All they need is the ability to track pallets at certain points in the supply chain, which will give them a good indication of global pallet movements.

RFID tag readers are also quite cheap so you could potentially fit a lot of them at relatively small expense. They could be embedded into the flooring of trucks, the entrances of warehouses and the arms of forklifts.

All of the examination of costs in the comments assume a zero cost of enforcement. Free courts, free police, free jailing or fine enforcement of any violators.

Without that free enforcement, these $5-$50 items would not be worth fighting over.

I don't believe that police protection should be a fee-for-service business, but are there no checks and balances on the use of public services for private gain?

Your argument applies to theft of any low value item. If a shopkeeper presses charges against a shoplifter, are they using public services for private gain? Should there be checks in place to stop shopkeepers from reporting too many thieves?

It would seem from the article that the business model not only requires the free protection, but that the enforcement of fines an judgements is a major driver of revenues.

In your hypothetical example, I believe you are referring to a physical storefront where the primary revenue generating activity is to sell merchandise.

If your hypothetical shopkeeper had a shop that made more money from forcing shoplifters to pay a fine for scaling, then... perhaps. I don't think your hypothetical shop exists, so it's tough for me to answer.

The problem with RFID for pallets is that the logistics space is still too decentralized. I work for a company which does a lot of custom RFID integration (we have worked with the companies in the article for tracking pallets.)

First, the RF properties of inserting RFID tags into wooden pallets just doesn't work well; it is far better to stick a label on the outside of the pallet. But then you get into durability issues, etc. So iGPS really had the right idea here, with molding the tags into the actual plastic pallet itself. Great read range.

The other problem is the hugely distributed nature of the supply chain. Pallet poolers like CHEP just don't have the resources to stick an RFID reader at every location that can possibly accept pallets, and distributors / store locations have no incentive to help CHEP out with returning accidentally-delivered CHEP pallets. RFID readers are unfortunately not cheap enough for CHEP to start parking readers everywhere to see where their pallets go.

Trust me, as a manufacturer of RFID readers, our company would absolutely love a scenario where RFID readers start being embedded in every trailer and on every forklift, however realistically that solution is still at least a few years out, and the cost needs to come down probably another order of magnitude to be practical for use in the pallet industry. iGPS was just too early to the game.

thanks for the insight! regarding embedding the tags I was imagining a small device around the size of a large coin with perhaps three prongs / nails attached. The prongs would embed into the wood keeping the tag in place, and the force of installation would depress the device into the wood slightly, but the RFID tag itself would remain on the exterior of the wood, so it should still be readable.

What if the pallets formed a mesh network? They would only need to have low-powered antennas (< 100m range) to network effectively with one another, and some kind of reader that talked to one could know all the other pallets that that one had "seen" and at what time. It would take a bit of work, and would be most practical for large pallet storage areas, but it could hugely cut costs, especially for wireless internet.

Reply on news.ycombinator.com