To Boldly Go …

As a long time space advocate (or perhaps in spite of being one), I was thrilled at President Bush’s announcement of a new venture to the Moon and eventually on to Mars with an integrated robotic/manned spaceflight architecture. I am hopeful in spite of prior efforts that have started grand and later turned to dust and ashes. I am hopeful in spite of the incredible torrent of articles and punditry saying that this new space venture will cost a trillion dollars and bankrupt the nation and will give Halliburton control of future drilling technology, or as a payoff to the military industrial complex who have not made enough from the war on terror.

I am hopeful in spite of articles by eminent scientists that it is science fiction to utilize the resources of the Moon to build rockets and provide fuel for the eventual trip to Mars or who say that we know everything about the Moon so why go there. Finally, I am hopeful in spite of those who think that it is an election year trick and will not go anywhere. The reason that I am hopeful is that… We can do this!

I don’t know how many people really and truly understand this, but we live in the year 2004. That is almost 40 years since the time of the design and production of the Apollo Command/Service Module and Lunar Excursion Module. When our forefathers went to the Moon in the dawn of the space age, it was with the computer equivalent of stone knives and bear skins. It is truly mind boggling to the modern computer engineer that we were able to go to the Moon with the hardware and software developed from scratch back then.

This is not to disparage but to honor those who made Apollo happen because in doing it, and in the aftermath of that program, many of those people went on to be a part of the revolution exemplified by Silicon Valley (my meaning here encompasses the whole American semiconductor and computer industry). Fast forward forty years and it looks like the engineering and technology from Silicon Valley has what it takes to make this new space era happen.

In 1969 the Lunar Excursion Module (LEM) sported the most advanced and compact computer built up to that time. It had 16k of RAM and a keyboard to type commands and instructions in machine code in real time. It had a small monitor to display data. It could control the navigation and flight of the LEM using radar data as well as transfer that telemetry in real time back to NASA on the Earth. It had a processing speed of more than a hundred thousand instructions per second. It was a remarkable computer in every respect. However… compared to now, it was so primitive!

Today a computer smaller, lighter, cheaper and using less power would execute tens of billions of instructions per second (that’s 10,000 times faster today not including multiple processor systems common today) and perform tens of billions of 64 bit floating point mathematical operations at the same time. That computer could input the global gravity map of the Moon, as well as a 3D radar terrain data overlain by a global 50 centimeter visual atlas of the Moon into its database and then perform all the mathematical calculations required to guide the spacecraft to a precision landing using input from real time laser altimeters, 3D stereo visual and radar sensors to match the real time with the predicted optimum trajectory and landing location. At the same time the communications system could transfer gigabits of data per second via a real time laser communications link to send back the landing in stunning high definition video.

If we are going on this great trek within the guidelines of the President’s budget, we will do it on the back of Silicon Valley, turning the technology wheel full circle from the day in 1963 when the first batch of small scale integrated circuits came off the production line at Texas Instruments. These chips were used to build the guidance system for the Minuteman missile (now used to launch small satellites for NASA and the Defense Department). There is a display in the Air and Space Museum in Washington DC that shows one of these early circuit boards.

Those who are electrical and computer engineers can see what the date codes are on the chips and know when they were produced. NASA and the Defense Department funded a huge amount of the nonrecurring engineering that was needed to qualify silicon chip production infrastructure and push forward silicon integration that forms the beginning of the curve now known as Moore’s law. Today that technology can help make exploration truly faster, cheaper and ……

The concept of “Faster Cheaper Better” for spaceflight projects was nobly conceived but often not implemented very well. The concept originated in Silicon Valley and is the heart and soul of the revolution that continues to shake the world and transform the way we live our lives. Most people today who have grown up with and use the Internet on a daily basis could scarcely think of giving up our access to the net. Millions of people are able to afford computers to connect to the Internet due to low cost; now less than $500.

Few people remember however, the early days of the Internet (or ARPAnet) when an Ethernet controller (the heart of all cable and DSL broadband connections) cost over a thousand dollars and the computer that hosted it cost over $5,000. Dial up modems cost hundreds of dollars and a disk drive had 10 megabytes that cost a thousand dollars. Even with the advances since then (that same controller now costs $30, the computer $400 and a 10 gigabyte hard drive costs $50 and holds thousands of high resolution pictures) profit margins remain and Silicon Valley, even after the dot com bust, is hugely larger than it was in 1983.

It would take NASA three years at its current $15 billion dollar budget before it would even spend the cash that Microsoft has in its bank account and that same amount would fund the first eight years of NASA’s proposed Lunar exploration effort. However, this is not all that Silicon Valley brings to the table.

“Faster Cheaper Better” failed because NASA tried to overlay that concept onto an existing aerospace industrial process. It was often said in the “Faster Cheaper Better” aerospace world, that you could pick any two, but you could not have all three. Silicon Valley does it every day. The ham radio community has done it as well many times by building satellites for a small fraction of the price of government contractor constructed units.

The British company, Surrey Satellite Limited, incorporates Silicon Valley practices and with that advantage, has the motto, we give you 80% of the capability at 20% of the price.” Silicon Valley lives in a world where a new technology can literally destroy old ways of doing business while at the same time reducing costs and improving productivity. How many who read this still use travel agents? Also, look at the music industry, absolutely in shambles by the rise of portable digital music devices, which made their old distribution system as obsolete as manual typewriters.

The aerospace world has ever so slowly integrated improvements in spaceflight hardware so that today a company can buy a super accurate micro machined integrated circuit chip weighing less than an ounce and uses a couple of watts of power to take the place of a expensive and complex set of inertial guidance components. Today a company can order off the shelf a flight qualified star field camera that weighs no more than a few pounds that integrates the Yale bright star catalog of over 20,000 visible stars with a microprocessor that can take an image of any section of the sky and match that with the catalog hundreds of times per second. Today a spacecraft can carry a flash memory hard drive that has gigabytes of onboard storage that also weighs no more than a few ounces that takes the place of heavy mechanical magnetic tape recorders.

Lets look at some other advances since the sixties. The solar array on the Skylab flight copy at the Air and Space Museum in Washington DC has 2 centimeter square solar cells that are eight percent efficient in converting sunlight into electrical energy. Today cells are available that are 4 X 8 centimeters or larger and are 28% efficient with custom cells that are over 30% efficient. Dr. Nevell Marzwell at NASA’s JPL led a group that demonstrated a cell at over 65% efficient as part of a recent space solar power effort, revolutionizing energy supply for spacecraft.

However, Congress and NASA defunded the project right at the point of success. Power supplies are dramatically smaller, lighter, and more efficient today. In 1968, the typical DC power supply on a spacecraft was about 70% efficient, meaning 30% of the energy was wasted as heat. Today anyone can buy DC power supplies that are typically 97% efficient. JPL has integrated Wind River’s VxWork’s real time operating system into the Spirit and Opportunity rovers to control the computers, sensors, and communications at a fraction of the cost of developing a system from scratch. However, for all of these advances the cost of spacecraft have not really declined much over the past 40 due to the NASA/contractor aerospace model.

In Silicon Valley the intense glare of global competition drives a corporate culture where the primary question is not why” as in the aerospace industry but why not”? Some might say, Peoples lives are at stake in human spaceflight. We can’t do it that way!” Have you driven a modern car lately? From the firing of the valves (eliminating mechanical cam shafts), the overall engine timing (eliminating the mechanical rotary distributor), to the electronics that monitors and adjusts engine for optimum performance and fuel economy in real time, you put your life in the hands of a computer every day.

The sensors in a car’s air bag have their origin in spaceflight to measure acceleration for inertial guidance systems. Real time computers and sensors such as this, multiplied by the millions, are on the road today. The automotive, semiconductor, and computer industry can serve as models of how reliability and production of silicon systems, woven together with robotics and human participants in the assembly process can be used to reduce the cost, improve the quality, and enable a Lunar/Mars development program within the budget proposed by the NASA administrator O’Keefe and the President.

For this process to work some legal method of working outside of the confines of the Federal Acquisition Regulations (FAR) and the NASA/contractor process and culture must be considered. The FAR and NASA processes are designed to monitor, control and cost the system developed by the aerospace industrial complex and the government over several decades. The FAR’s are completely inadequate to the task of producing the lowest cost, highest reliability, and fastest implementation for the Bush plan or any other plan for that matter. There are several reasons for this.

One reason is that profits are capped at a number around 8% of the total contract value. When the government buys a copy of Microsoft Office or Windows, they do not tell Microsoft what their percentage profit will be. Secondly, there is a culture that has evolved (different than the one that the CAIB commission addressed) associated with the contract process. NASA or the Defense Department have proposal preparation instructions with guidelines associated with manpower loading, expected overhead rates, and projected costs for development. This has led to a twofold problem. A company has as its prime directive, the maximization of profits and business opportunities.

Therefore, in order to get business, companies often underbid contracts because to win the contract. After winning the contract, the way to maximize profits is to maximize the overall costs. This is evident in almost every NASA and Defense Department contract today. The only way out of this quagmire is to develop new ways of rewarding companies that want to participate in the exploration initiative without becoming part of the contractor culture.

Silicon Valley can be the enabling factor that can bring the full weight of modern technology and management style into the space world and the Bush initiative to revitalize a system grown old and inefficient. Very few Silicon Valley style companies want to bid on government contracts for the reasons enumerated above and more. The government is often considered a bad customer in the Silicon Valley world with late payments, delayed paperwork and a contract award to payment cycle that can take six months to execute. Add to this, the capricious natures of congress and NASA that award contracts and then take them away, or change the whole program focus to eliminate contracts in progress.

For most high tech commercial products, the entire product life span is less than the development time in government development contracts. The contractor culture only works because aerospace companies have adjusted their corporate culture and personnel into this archaic mold. New entrants have to adjust to this culture or lose out on contracts. So the key to bringing in the Silicon Valley culture without destroying it is to develop ways to reward these companies for performance, not effort.

For this to happen the government must set aside a portion of the proposed NASA budget to fund specific items outside of the bidding process. This is called a Contingency Contract Arrangement (CCA). A CCA is set up in the following way: NASA wants to open up the exploration of the Moon and obtain scientific data. There is a plethora of science missions that have to happen as a precursor to maximize science return. For example we need a high resolution visible map of the Moon. This could be at a 50 centimeter (approximately 19 inch) resolution, entirely adequate to locate and identify dangerous terrain that might crack up a landing mission.

What could be done, since it is the consensus of most of the science and exploration community that this data is needed, is to provide the funding in an account to purchase this information. Lets say that NASA puts up $90 million dollars on the visual map and $100 million for a 3 meter resolution radar map. The government would only pay for the first company that successfully does this mapping with nothing for second place. The government would also put a strict time limit on the offer, say 24 months to get to the Moon and another 9 months for the data. If a company is one day late it is out of luck. The observer might say that there is no way any aerospace company would do this. Exactly the point! However, Silicon Valley lives with this level of risk every day and there are both companies and entrepreneurs who would be willing to go for such a deal.

What the government loses is control of the process. This is unimportant as long as the product is what the government wants at a price that they are willing to pay in an acceptable time frame. What the NASA peer review community would do is to determine the exact data set that they need, the format that they need it in and the delivery method. If the company that provides the data, in the right format, delivered how NASA wants, and within the time period specified, is able to do all of the above at a price below that of what the government has set as the price then great! That is called profit.

Normally, if NASA contracted for this, the company building the imagine spacecraft might make $7.2 million dollars (at a 8% profit rate) but if there is an overrun, NASA usually pays that as well, along with the profit on that overrun. In the new scenario, if it costs the company $200 million to provide the data, they get $90 million dollars, or if it cost them $40 million dollars, they get $90 million dollars. The government gets no oversight role other than to verify the format of the data and its quality. This is simpler for the government and simpler for the provider of that data. If the money is not claimed within the required time period, the CCA is voided and NASA keeps the money to do it the traditional way. This is a win-win scenario for the government and data providers.

If the first two of these CCA’s are successful then the concept could be extended to cover landing and even manned missions associated with the exploration effort. The benefit here is that private enterprise could bring a new and refreshing burst of life and technology to the space development effort.

Today the contractor world is very conservative by nature and is reluctant to bring new technologies to bear on spaceflight efforts because of the fear of embarrassment if something fails. Private enterprise can be conservative as well but there is a culture of measured risk/reward that is absent in the government/contractor world. If this new contracting model works there will be some revolutionary new spacecraft designs that could completely transform the state of the art. There would be new entrants into the field of spacecraft development bringing welcome competition to a defense industry that has seen the hundreds of companies that existed in the sixties turn into just a few megagiants with a few scrap feeders on the sidelines. This process would also bring new engineering talent into the field.

The end result is to be able to implement the President’s vision for the exploration and development of the Moon and to be able to use the resources of the Moon to go on to Mars and to benefit all humanity here on the good Earth.