Dell just announced the largest recall of any consumer electronic item in history: 4.1 million lithium-ion batteries manufactured by Sony and used in over 30 of its models. The recall was precipitated by a small number of overheated batteries in notebooks getting hot enough to cause a fire or potentially explode. Although the risk of such mishaps is small, Dell did the right thing for its customers and itself by taking no chances on someone getting hurt (or worse).

Obviously, a recall of such magnitude will not be inexpensive. (We estimate that it will cost Dell at least US$50 for each battery, or over US$200 million, although some substantial amount will be funded by Sony.)

But this recall is only the tip of the iceberg and hardly an isolated case. I believe the risks of overheating in modern portable devices is increasing as more densely packed devices and higher charge rates become the norm. Companies and users must be aware of this risk and do whatever they can to protect themselves and their property from potential heat-caused accidents or suffer the consequences.

This is not the first time batteries have had overheating problems or been recalled by a computer vendor. Dell has done it before, as have Hewlett-Packard, Apple Computer and others. And while vendors quickly rallied this time to say this is not a widespread industry issue, that is not the case. It is fair to ask whether this is something unusual or whether it will become an industry trend as more battery-powered devices make their way to market.

Most manufacturers of modern battery-operated devices face a dilemma. Users expect these devices to run for hours, even as they become increasingly powerful and power-hungry. They want their notebooks to run a minimum of four to five hours, up substantially from the two-to-three-hour average of just a couple of years ago. But battery technology hasn't kept pace with the "silicon curve." Indeed, battery power gains only about 5 percent to 10 percent per year; a far cry from the high multiples that processor performance increases by every 18 to 24 months. Users want ever smaller and lighter devices, but batteries are a major portion of the overall size and weight of nearly all portable devices, especially notebooks.

Furthermore, increasingly impatient users (myself included) want fast charge times that allow them to get up and running quickly. That takes chargers that pump high levels of power into the battery, thereby increasing heating and limiting the battery's lifetime, which at best is generally only 250 to 300 charge cycles. Some new battery technologies like fuel cells may change all this, but don't expect such technologies to be made economically and technologically viable for the mass market for at least three to five years.

Designers of portable systems (not just notebooks, but things like handhelds, mobile phones and music players.) must make trade-offs between device size and weight, battery life, charge times, power supply size/weight, overall power density/heat dissipation and, of course, cost. And it's not just batteries that are at risk of overheating. Power supplies (the little "brick" chargers) have been known to pose fire risks as well. Because many notebooks are resold from third-party manufacturers in the Far East, it is not uncommon for marketers such as Dell to allow the reseller to design much of the notebook. Furthermore, marketers may not always adequately test the devices they sell under their brand names. This is especially true for smaller brands and devices targeted at consumers.

With all of these potential problems, what's to be done? User companies should be cautious about the portable devices they purchase and deploy, even from the major enterprise marketers like Dell, HP and Lenovo Group. Few companies evaluating purchases of portable devices (not only notebooks, but also smart-phone devices) do an adequate power/temperature/life-cycle test. I believe that, based on the increasing density and the higher charge rates inherent is so many of the new devices, companies must start evaluating the overheating risk of these devices. In fact, companies should require suppliers to provide them with heat-testing data to assure that no risks exist and that the suppliers have actually tested the devices adequately (most have not). Furthermore, companies purchasing devices for employees should check with their insurers to confirm that they are covered for any liabilities in case devices overheat and cause burns to the end users or start fires. Finally, enterprises and device vendors should educate users about the inherent risks associated with the literally hot new devices making it to market.