"He giveth His beloved in their sleep" (Psalm 127:2)

 ø Finnish mathematician Antti Lange has invented a new computational method.

- The incredible thing is that thousands of mathematicians have for decades been searching for a fast Kalman Filter (FKF) but they were not able to discover it. It is also amazing that I found it through a dream after having received a promise from a Bible verse, tells Antti Lange.

 ø There has been debate in recent years that mathematicians who obtain patent rights for important formulae or so-called algorithms may get a griphold over society in the USA, the first country to issue patents for computational methods. The successful patentees for formulae with great industrial applicability may become the world's wealthiest people. Finnish pentecostalist Antti A. Lange could well become one of them. He is a long standing member and former parliamentary candidate of the SKL (Finnish Christian Party) and has been working for the meteorological office in Helsinki, Finland.

Experience from youth

- In 1967 I was serving national service in the general staff mathematical research offices. As I was a conscientious objector a meteorological problem I was assigned to me. I failed to solved it satisfactorily with existing computers of those times. I began to work hard with the formulae themselves. When I was inverting so-called "matrices" the computations were spread out on papers all over the place. An experienced mathematician would have abandoned the idea, but for some reason I desired to render a report to my kind superiors. I would not give up. Finally, the calculations started to contract and everything condensed into a surprisingly simple formula. This experience came to mind later in my dream.

In a super computer centre

- I entered the metereological office in 1969. I had previously worked as a senior mathematician in the University of Jyvaskyla, at its computing centre.

- Following Finland's engagement in a super computer centre near London for medium range weather forecasting I began to make regular visits there. Later I was attached to it as a consultant for a four months period in 1984.

- I had a difficult problem to be solved: How to make satellite, radiosondes and other weather observations commensurable in a practical way? There are at least as many standards of measurement in the world as there are nations. This causes a serious problem in weather forecasting as every country has instruments biased in its own direction.

- At that time I transacted poorly in the international environment using the English language. If one was to take on a task for consultation with only a short limited amount of time, without the ability to report on everything attempted, in practicality it means the severing of one's own career. In such cases I prayed and worked morning til night.

Dream following Biblical promise

- Suddenly a phrase from the Bible bounced out: "It is vain for you to rise up early, to sit up late...For so he giveth his beloved in sleep" (Ps. 127:2). I was tempted to think the Lord was mocking me in my stress, until I understood that this was a promise.

- One night I awoke with tremendous joy. In a dream I had solved the problem. In this dream the Holy Spirit spoke to my inner being and took me back to my army situation. I was advised to replace one matrice with another in my old simple formula, and the solution was obviously there:

- The outcome of this was the extension of my consultancy contract. I worked in the super computer centre in England for four years in all. 

Return from London empty handed

- While in England we accomplished a super computer test in which a clear improvement in weather forecasting was demonstrated through the solution I had furnished. However, even though one could demonstrate improved weather forecasting, as the Centre serves the whole world, no vague aspect could be attached to it.

- In a super computer environment a vast amount of calculations are being carried out daily. We computer mathematicians know how severe the problem of the so-called "numerical instability" of these calculations can be. In other words, during this incredible amount of computations, errors in rounding off may start slowly piling up and clogging, mixing up the final result and jamming the end sum. This only needs to occur in one part of a huge super computer system and it will cause the collapse of the whole forecasting process.

- A rigorous proving of the numerical stability of this God given invention turned out to be an enormous difficulty. I simply could not do it. I had to leave London as a failure in 1988. 


- On return to Finland I received no encouragement from any rival computing expert or researcher, though I did from the management. A Finn brings from abroad either promotion or a grave for his career. I was of the latter sort. Due to these difficulties I decided to throw the whole idea into the rubbish bin, because it only seemed to cause problems. However, we have prayer meetings at home on Mondays and I decided to yet again bring the matter before the Lord. Had I been misunderstanding these matters or was the solution of great value after all?

- God answered my prayer immediately on the following day. Mr. Raoul Laine, a missionary from Thailand had arrived for a furlough in Finland and told of a revelation during our church meeting: "a person had received a precious pearl as a gift. This pearl has shattered in his hands and he is depressed". But then in this revelation the person was told: "Repent, and I will give you a new one." I was dominated by bitterness towards my colleagues, because they refused to understand what in my understanding was a clear case and had left me alone. I realized that I had to surrender my bitterness and not give up.

The same results as the Kalman Filter (KF)

- Within a couple of months my problem was solved: I found that the required numerical stability was dependent on certain conditions. I noticed that these conditions are exactly the same as with another computational method called the Kalman Filter. If my formula is employed exactly as a Kalman Filter then all the important results that Professor R. E. Kalman had already verified at the end of the 1950's apply.

- The Kalman Filter was used for the first time during USA moon space projects and the method has now spread extensively. In fact, every automatic navigation device makes use of a Kalman Filter.

- The first "pearl" which I believe I received from the Lord was obviously limited to meteorological observing systems. The new "pearl" covers almost all automatic process control.

WIPO recognition

- When I managed to derive the exact connection between my own computational method and the Kalman Filter, the clogging problem was rigorously solved. I sent my invention to the patent bureau in 1989.

- It passed through the investigations and screening processes of the World Intellectual Property Organization (WIPO). In 1991, I received the required favourable statement without reserve from WIPO. This testified that my invention had never been discovered by anybody else previously; that it contained an inventive step; that it works and is industrially applicable. I hastily forwarded patent applications to all countries involved in the Patent Cooperation Treaty (PCT) of the world, tells Antti Lange.

- It is indeed a simple mathematical formula. It isn't much stranger looking than the theorem of Pythagoras, but that is where its strength really lies.

"These formulae are new and because of their great use obviously surprising", says the worldwide patent organization's statement.

Wide scope of application

- I am applying for a method patent relating to precision measuring. The patent application is claiming sole rights for the use of my computational method relating to: all calibration devices which need real-time precision regulation; automatic navigation devices; weather condition determination or forecasting; geophysical, social or economic prediction formulation; industrial control systems; fire and theft protection automative equipment, etc. In a way this could be compared to the theorem of Pythagoras. Where would it not be needed in one form or another?

One computer replaces a thousand

- A year ago an authoritative American review panel reported that the Kalman Filter is the only rigorous way of making weather forecasts scientifically. The same panel complained that it is computationally impossible for the Kalman Filter to be applied to these large problems. However, they did hope to see light at the other end of the tunnel, and concluded that perhaps sometime in the future it would be possible with the aid of a thousand super computers. We realize how expensive that would be! In Finland there was some anguish in purchasing just one super computer.

- Now I have used to advantage a young gifted Estonian mathematician. According to a very preliminary assessment only one super computer might suffice when employing my invention method (FKF).

- By employing my FKF method the majority of large Kalman Filter problems are speeded up a hundred times, and in weather forecasting possibly by as much as a million. My invention, however, does not include any approximations. It is an exact solution.

Super believer or..?

To the question as to whether Antti Lange is a super believer higher than others he replies:

- A Bible verse spoke to me that Elijah was subject to like passions as we are. I recognize my own wretchedness and weakness. Nevertheless I desire people to believe the best about God. He has not created anyone unnecessarily. To Him each person is a favorite child.

- It is extremely dangerous to argue about being a friend of God when one is sinful flesh. As a word of warning and to bring us down to earth are Jesus' words to Judas Iscariot: "Friend, why have you come?" God shows His glory as great grace or judgement to those that are close to Him, thinks Antti Lange.

August 15, 1991


(translated by Allan Scotson from a news article published in Kristityn Vastuu.)

Link to FKF.net.
* Last revised: June 23, 2000

Bach: Kb Conc No5 in F- BWV1056 2Largo
Sequence © Pierre R. Schwob - by permission. Original from the Classical MIDI Archives