Robot Security Guards

It has been a while since I’ve discussed new security technologies, but an article at TechnologyReview.com peaked my interest this morning. It’s an article about a robotic security guard currently being deployed in Silicon Valley. The company web site makes it pretty clear that the current model is not really designed for a chemical manufacturing environment, but it does provoke some interesting ideas.

Why Robotic Security Guards

Ninety-nine percent (hopefully) of security guard work is boring, repetitive and low skilled. This is one of the reasons that it is generally a low pay job. As with other areas of commerce it makes a certain amount of sense to replace boring, repetitive tasks with technology that can do it effectively at lower cost. There are two areas of security guard operations that may be ideally suited to robotic replacement; routine patrols and alarm response.

Patrolling a large complex area is one of the most boring tasks that security professional has to deal with. To be effective patrols have to be both random and complete. Security guards tend to establish routine routes that are both predictable and avoid unpleasant areas. Trying to enforce randomness and complete coverage requires frequent supervisor checks which are generally expensive and ineffective.

In large complex environments like chemical manufacturing facilities, the use of sensor technology (including video and intrusion detection devices) has become an increasingly favored method of reducing staffing requirements for security guards. Unfortunately, the more sensitive these systems become the more often there are false alarms.

The problem is that every alarm, false or positive, must be investigated to determine which it is. The larger the number of false alarms that must be investigated the larger number of security guards you need to have on hand to do the investigation. You can reduce the ratio of false alarms to positive alarms by reducing the sensitivity of the sensor network, but that makes the network easier to penetrate.

What Would a ChemSecBot Look Like?

You need a mobile platform with onboard video, navigation and communications technology. The basic platform described on the company web site would probably suffice with an upgraded maneuver system designed to operate in a more rugged (including maybe road) environment. Since night lighting is not always adequate within all areas of a chemical facility perimeter, some sort of night vision system would also be a good thing to have.

Sensor technology will be the area that differentiates a ChemSecBot from their more normal security counterparts. At a chemical manufacturing facility one of the things that everyone (or every bot) ought to be watching for 100% of the time is chemical leaks; particularly hazardous material leaks. Each chemical facility would require its own unique set of chemical detector sensors, depending on which chemicals are in use at the facility.

Almost every chemical facility has flammable atmospheres present in part of the facility from time to time; even if it is only in the event of a spill. Any ChemSecBot built to operate in a chemical manufacturing environment will have to be certified to safely operate in a flammable environment.

Since it is unlikely that the ChemSecBot would be able to climb the stairs and ladders found in a chemical manufacturing facility, it would also be helpful if it had the capability to raise a sensor array above ground level.

Emergency Response Uses

Since, based on current history, a chemical facility is more likely to experience a chemical spill than a terrorist attack, it would be helpful if the ChemSecBot could also have an emergency response role. If the interior workings of the system could be protected from exposure to the hazardous chemicals seen in a particular facility, one could easily envision a reconnaissance and observation role for the Bot. The inclusion of appropriate chemical sensors able to measure concentration would be an invaluable emergency response resource.

Moving Forward

While we are a significant distance in time and technology from deploying a fleet of ChemSecBots, it does seem to me that industry would be willing to pay more than the $6.00 an hour discussed in the article I mentioned at the start of this post. While there may be a larger unit market for MallBots, the chemical security industry will probably be a more important market segment in the long run.