If you're a CXO overseeing a critical infrastructure that contains SCADA (supervisory control and data acquisition) controls, a chief concern is how to protect the infrastructure against terrorist attacks. Changes in control software will continue to accelerate until the most critical infrastructure weaknesses (oil refineries, electrical power plants, water treatment facilities) are addressed worldwide. But it may take years to replace all of the controls.
In order to address some of these concerns, networking vendors are deploying solutions to monitor network traffic between the management systems of these controls to determine the validity of its state. They can plan on implementing authentication and access controls on the sessions that communicate with the controls. As a newer generation of controls is deployed, authentication and authorization features will be built into the controls themselves. All access can be logged to determine if there is any tampering.
But there are other concerns associated with the electrical power, refinery, chemical plant, water treatment, and nuclear power industries that deploy these new controls. There will be a great need for custom simulation software for specific vertical industries like the ones listed above. Simulations will be needed to determine what will happen if a new set of policies (control states) are implemented.
The control settings (policies) need comprehensive testing. It will be too difficult to determine all of the various states of the controls and their interaction with other controls via spreadsheets. The dangers could be catastrophic, such as chemical or waste spills, so the software will need to be very sophisticated to manage the various good and bad control permutations.
This SCADA simulation software reminds me of the live/dead analysis that goes on within Energy Management Systems used by electrical power companies to manage their multi-state electrical grids. Live/dead analysis simulates the response to an electrical line change to a portion of the grid. The change can then be implemented if the simulation shows it is safe.
What are the difficulties associated with this new simulation software? It will be difficult to create software that properly models control systems, such as an oil refinery. It is likely that the software will be customized for different corporation's refineries. These customizations will need thorough testing before the simulation software is fully deployed. Software errors in the simulation software could also lead to disasters so the software may need to pass a certification process before being deployed in a refinery's private network.
Other concerns include the internal network's wireless connectivity to these controls. I'm concerned about the wireless connectivity between the simulation system, the live network, and the controls. This may require a hardened and/or specialized wireless network designed only for communication with critical infrastructure systems.
In the next few years, critical industries will shore up their SCADA control weaknesses. There will likely be standard software for different vertical industries, like oil refineries, which is then customized for a given company. There will be a need for better simulation testing due to having more complex controls. This core complex software (without the customizations) for each vertical industry may need to be certified for safety reasons. Lastly, an intelligent collection/alteration of SCADA control data may require specialized, hardened, wireless communications for protection from terrorists.
This story, "The future of SCADA-control security" was originally published by CSO.