It is assumed that targeting inspection system on PSC is based on the statistical particulars of PSC inspections in the region. The aim of the current research is attempt to define the target factor (TF) for selection ships for an inspection as objective system basing only on the data contained in the database and reflecting current data on every moment.

 The main idea of the proposed approach is calculation of a figure on a ship (overall target factor - OTF) as a combination of elements (L) produced from the ship inspection history and compared with the average values of the same elements for all ships in the database.

 To implement the above approach first of all it necessary to recognize dependencies of OTF elements from the various data contained in the database. It is assumed that for the initial (basic) data the ship particulars and inspection results are accepted. There are:

 Some other data may be added to the list above subject for more detailed analysis.

　

 Description of data mentioned

 Data related to time.

 Ship age, and time from the last inspection are time-related data. Actually, in the database the only year of build is recorded for each vessel. So, the ship age may be calculated as a difference between current date (year) and the year of build of the ship. Since exact date of build of a ship is unknown the number of years of age is considered. Time from the last inspection is a timeframe between current date and the date of the last inspection of the ship.

 Data related to lists.

 Ship flag, ship type, ship CS, and Authority of the last inspection may not be reflected in numbering system directly. At the same time the variables exactly determined and each ship is definitely belongs to exact value.

 Data related to numeric characteristics.

 The numeric data are originally stored in the numeric format and may be used in that format. At the same time it is recognized that the absolute values of numeric date are difficult to compare. Therefore it is proposed to consider relative values instead of absolute ones:

　

 It is recognized from the current set of data collected by the APCIS database that the data may be divided onto two groups:

 Therefore the OTF may be calculated as

　

　

 Group Factor is a combination of elements related to ship groups (or better to say to ship particulars) independently on results of inspections of the particular ship. The GF may be defined as:

　

　

where i- index of an element,

 Taking into account structure of the inspection report (Forms A and B) and set of data presented in the APCIS central database the following elements of L^GFi may be represented:

 And, thus, domain of i in (2) is {t, f a, RO}.

where Xdef - ratio of inspections with deficiencies,

　

 On the initial stage of the developments it is suggested to use a simplified function for L^GF:

　

where px - weighting coefficient for an appropriate parameter,

 The simplest way for the px identification is to define it equal for all parameters. But, a more effective approach is calculation of the px from the current inspection data. Taking into account the aim of the TF system, that is, selection ships for inspection from all ships available, it is logical to assume that the px should depend on a range of values of appropriate parameter for inspection records collected. If so, the value of the px may be defined as:

　

where sx - standard deviation of parameter x for a sample X,

 The are two methods may be used for definition of parameters X in (4) - direct (absolute) and reduced ones. For the reduced method value of Mx is recognized as 0, and, therefore, a value of X < Mx leads to negative correspondent reduced value X^*. The negative value on its turn reduces overall value. One positive thing for the reduced method is dimensionless variables. For the reduced method (4) may be redesigned as:

　

　

rДe X^*idef = (Xⁱdef - Mⁱdef) / Mⁱdef,

　

 Decompressing formula (2) the following Group Factor is a result:

　

　

 Weighting coefficients k^GFi define influence of a particular factor to final value of GF. Two methods of definition of the coefficients may be considered. The first one is expert method where a value is stipulated by experts (the method is used in the current Tokyo MOU TF system). The second one is statistical method. For the statistical method values of coefficients are calculated as:

　

　

where i - factor index with domain i = {a, t, f, RO};

　

 Historical Factor (HF) establishes dependence of TF from the results of the previous inspections of the ship.

 Similarly to the definition of GF the HF may be introduced as:

　

　

where k^HFi - weighting coefficient of element i;

　

 Considering data definitions given above the L^HFi is defined as:

　

where qY = sY / qMY - weighting coefficients of appropriate arguments,

 Redesigning (10) to dimensionless mode, the following formula is received:

　

where Y^*idef = (Yⁱdef - Mdef) / Mdef;

　

 And finally for HF:

　

and

　

　

where i - index with domain i = {def, det, adef adet, nr, A, τ};

　

 In conclusion it should be mentioned that the above is just first attempt to develop objective approach to TS system independent on political, or regional trends. The method requires some more detailed developments including more careful definition of variables, elements and parameters of the system. Some testing on actual data should also be performed.