IBM UBIQ

Low Bandwidth Visual Communication

The UBIQ system sends video from a remote field location via a mobile phone to any computer connected to the Internet. The system is more interactive than simply sending a collection of still images via email from the phone. It also has a selection of low bandwidth "viewfinder" codecs to allow a remote expert to guide the phone user to the correct position for snapping a high-resolution picture.

UBIQ Block Diagram

What is it good for?

  • inspection - construction projects
  • troubleshooting - copier repair
  • consulting - remote clinics
    

Advantages

  • reduced travel costs
  • improved response time
  • leveraged training
  • fewer restrictions
    - personnel selection, site accessibility

Field Inspection

There are a number of scenarios in which UBIQ would be useful. One example is field inspection in the construction industry. Suppose a US company was building a large hydroelectric dam in Brazil. It is standard practice to have at least one engineer on-site at all times to ensure that appropriate materials and techniques are being used. Then there are periodic inspection visits from the home office to double-check the project and garner overall progress information. Finally, sometimes things go wrong, like a concrete slab starts slipping down a hill, and an expert must be flown in. Typically, most work at the site stops until this individual arrives and diagnoses the problem. This can mean a lot of idled manpower and machinery, not to mention schedule slippage which can have significant economic ramifications. It is even worse if the company sends a civil engineer to straighten out the situation, but it turns out a hydrologist is needed instead. This adds extra transportation costs and work disruptions as the correct individual is first located then mobilized.

Now imagine that the company employed UBIQ. First, instead of conducting site inspections once every two months, the company could "beam-in" their own chief engineer every week at very little cost. Any worker already on-site could be drafted into carrying (or wearing) the transmitter around the work area -- no specialized skills are required. With closer supervision, there is less chance of a major problem occurring. Furthermore, there is no need to obtain foreign visas; no delay or expense in arranging flights, hotels, and ground transportation; and no human resource problem with engineers complaining they are away from their families too much. Also, should some unforeseen event occur, an engineer can be beamed in right away without even waiting for the next scheduled airline flight. This also allows timely intelligent intervention when physical travel is outright impossible, such as to offshore oil platform during a storm. Moreover, if the wrong personnel are originally sent, it is easy and quick to fix the mismatch -- just have the correct specialist connect to the feed instead. Again, there is no delay for travel or visa clearance which means less costly downtime on the site. While the example has been cast in terms of international operations, UBIQ can even benefit more local endeavors. Why should an engineer have to spend 2 hours in q car to cross from one side of the city to the other during rush hour just to render a 10 minute judgment? While not every problem can be resolved without true physical presence, a substantial fraction are likely to be amenable to this technology.

Maintenance

Now consider a more routine field-service and maintenance scenario involving photocopier repair. Businesses really need their machines working all the time and will often pay for high priced service contracts that guarantee a technician will be at their facility within an hour. With a large territory to cover and the vagaries of urban transportation, this typically means the service provider must recruit a large staff of field service personnel. Moreover, the customer really wants his machine fixed, not just looked at. This means each field technician must have a lot of training so he can fix even the most difficult problems. Such training is expensive and becomes more so if there is a high rate of employee turnover or if the product line changes frequently. And not every individual is able to learn the details of the system deeply enough to be useful, so a premium must be paid to lure the more competent applicants. Furthermore, the system is inefficient is its use of all this training. Field service personnel spend a large fraction of their time driving from site to site, not even using their expertise. Other times, they sit around waiting for a call which is not only a waste of skill but can also be very boring and lead to a loss of employees.

With UBIQ a number of these problems can be solved. First, 90% of the time the problem is very simple, such as the machine is unplugged or is out of toner. It does not take extensive training to notice and fix such difficulties -- it is the other 10% of the situations which require more expertise. Thus the service provider could instead take people with minimal training and have them cope with these easy problems. When they found something they did not understand they would contact the home office and request an expert be "beamed in". With sufficient video information and a semi-skilled set of "remote hands" there is a good chance that even difficult problems could be fixed. Moreover, fewer true experts would be required. They are only needed for the tough cases and their time could be utilized more effectively since they no longer spend any of it in travel. Also, a load balancing process could redirect calls to whatever experts were currently available, irrespective of physical location. Furthermore, it should be easier to retain the experts in whom the company has invested a lot of training. They can live wherever they choose (the mountains in Utah?) and they get only the interesting problems, not the humdrum ones. In fact, part of the training of new experts might be alleviated by the "caching" of special techniques by the personnel currently out in the field. Every time they consult an expert on how to solve a particular problem they essentially receive just-in-time on-the-job training which may allow them to handle a larger percentage of the cases without intervention (or eventually fill the role of remote expert themselves without as much explicit instruction). With UBIQ the result is the still same for the customer -- quick repair -- yet the operating costs for the service provider are now lower. These savings could be retained, or alternatively shared with the customer to yield a competitive advantage for the provider.

Medical Applications

Similarly, UBIQ could be useful in medical applications. Imagine there is a clinic in Botswana. The local doctor is probably a general practitioner and can successfully treat common ailments such as foot infections. However, he may have trouble with more complicated diagnoses like various forms of nerve damage. Most likely he will not have access to sophisticated lab analysis or advanced medical imaging, and it is unlikely that his patients can afford to travel where the appropriate specialists are (or even get a visa out of the country). It is equally unlikely that a high-paid specialist will take the time and expense to visit the foreign clinic, especially in a timely manner. Moreover, there may overwhelming reasons for not going in person, like fear of contagion if an epidemic is raging. However, just "beaming in" to have a look at a patient is much less of a chore. For this sort of consultation a phone call often is not sufficient. The doctor might need to look at the color, texture, and extent of a rash, or might need to see the range of flexibility in a joint or exactly what sort of tremors the patient's hand exhibits. Once again, the same technology could also be used on a much more local scale, where a doctor in New York brings in a "virtual specialist" from New Haven. In some sense, the human body is as much a machine to be maintained as a copier is.


Media

  • Overview - Presentation explains the motivation for building this device and some of the benefits of using it (1MB PPT).
  • Video demo - Shows the interaction of the PC and phone and how to perform certain operations (10MB WMV, audio commentary).
  • Presentation - Longer presentation contains more technical details about the various codecs and their implementation (3MB PPT).
  • Codec comparison - Shows the output of 3 different low-bandwidth codecs on the same video sequence (11MB WMV, no audio).

Downloads

  • Quickstart guide - Shows how to set up and use the prototype.
  • Windows client - Works under XP to get images from the phone (see video above).
  • Phone application - Used for re-launching the server in photo mode.
  • Phone server - This is in Windows Mobile 5.0 for the ARM V4I processor and is known to work on the Verizon VX6700 (aka Audiovox 6700 aka HTC Apache).
  • Windows server - Alternative for use on a PC with an attached webcam.


Contact: Jonathan Connell (jconnell@us.ibm.com)

Last updated: 1/22/08