REMOTE MONITORING APPARATUS, REMOTE MONITORING MAINTENANCE SYSTEM, REMOTE MONITORING METHOD, AND REMOTE MONITORING PROGRAM Technical Field [0001] The present invention relates to a remote monitoring apparatus, a remote monitoring maintenance system, a remote monitoring method, and a remote monitoring program to calculate a failure probability of a part included in a monitoring target device. Background Art [0002] A conventional remote monitoring maintenance system calculates a future failure probability from an elapsed period of use and the like of a part included in a monitoring target device, and calculates a maintenance cost of each part based on the calculated failure probability. Based on this maintenance cost, a maintenance worker determines an inspection time to inspect the monitoring target device (for example, Patent Literature 1). Citation List Patent Literature [0003] Patent Literature 1: JP 2009-217718 A Summary of Invention Technical Problem [0004] A technique described in Patent Literature 1 does not calculate a failure probability based on information (hereinafter referred to as maintenance information) obtained by the maintenance worker upon visiting a site and inspecting the part of the monitoring target device. Generally, the failure probability predicted from the maintenance information obtained by the maintenance worker upon an inspection is more accurate than a failure probability predicted from device information (for example, information on an elapsed time of use and the like of each part) transmitted from the monitoring target device. However, if the maintenance worker checks all of parts at each inspection, it is costly and time-consuming, and it is not possible to efficiently perform an inspection work. On the other hand, if the inspection time of the monitoring target device is determined based on the failure probability calculated only with the device information, an unnecessary inspection work occurs or the like so that it is not possible to efficiently perform the inspection work. [0005] A remote monitoring apparatus, a remote monitoring maintenance system, a remote monitoring method, and a remote monitoring program according to the present invention aim to solve a problem described above, and aim to calculate a failure probability using device information transmitted from a monitoring target device and maintenance information obtained by a maintenance worker upon an inspection. Solution to Problem [0006] A remote monitoring apparatus of the present invention includes: a failure probability calculation unit to calculate a failure probability for each part of a monitoring target device at a next inspection, based on device information acquired from the monitoring target device; a failure probability correction unit to correct the failure probability based on maintenance information at a previous inspection by a maintenance worker who inspects the monitoring target device; and a failure risk calculation unit to calculate a failure risk at the next inspection based on the failure probability corrected by the failure probability correction unit. [0007] A remote monitoring maintenance system of the present invention includes: a device information storage unit to sore device information acquired from a monitoring target device; a maintenance information storage unit to store maintenance information at a previous inspection by a maintenance worker who inspects the monitoring target device; a failure probability calculation unit to calculate a failure probability for each part of the monitoring target device at a next inspection, based on the device information; a failure probability correction unit to correct the failure probability based on the maintenance information; a failure risk calculation unit to calculate a failure risk based on the failure probability corrected by the failure probability correction unit; and a maintenance item selection unit to select, among parts of the monitoring target device, a maintenance item of a part whose failure risk calculated by the failure risk calculation unit exceeds a threshold value, as a priority maintenance item requiring an inspection at the next inspection. [0008] A remote monitoring method of the present invention includes: a failure probability calculation step to calculate a failure probability for each part at a next inspection, based on device information acquired from the monitoring target device; a failure probability correction step to correct the failure probability based on maintenance information at a previous inspection by a maintenance worker who inspects the monitoring target device; and a failure risk calculation step to calculate a failure risk based on the failure probability corrected by the failure probability correction step. [0009] A remote monitoring program of the present invention causes a computer to execute: a failure probability calculation step to calculate a failure probability for each part of a monitoring target device at a next inspection, based on device information acquired from the monitoring target device; a failure probability correction step to correct the failure probability based on maintenance information at a previous inspection by a maintenance worker who inspects the monitoring target device; and a failure risk calculation step to calculate a failure risk at the next inspection based on the failure probability corrected by the failure probability correction step. Advantageous Effects of Invention [0010] Since a remote monitoring apparatus, a remote monitoring maintenance system, a remote monitoring method, and a remote monitoring program according to the present invention calculate a failure probability based on maintenance information in addition to device information, it is possible to calculate a failure probability more accurate than a failure probability calculated only with the device information. Therefore, it is possible to determine an accurate inspection time and to efficiently perform an inspection work. Brief Description of Drawings [0011] Fig. 1 is a configuration diagram of a remote monitoring maintenance system according to a first embodiment. Fig. 2 is a functional block diagram of a failure risk processing unit according to the first embodiment. Fig. 3 is an operation flowchart of the remote monitoring maintenance system according to the first embodiment. Fig. 4 is an operation flowchart in which a remote monitoring apparatus according to the first embodiment calculates a failure probability. Fig. 5 is a diagram explaining an aspect of implementation of the remote monitoring maintenance system according to the first embodiment. ■ Fig. 6 is a configuration diagram of a remote monitoring maintenance system according to a second embodiment. Fig. 7 is an operation flowchart of the remote monitoring maintenance system according to the second embodiment. Fig 8 is a diagram explaining an aspect of implementation of the remote monitoring maintenance system according to the second embodiment. Fig. 9 is a configuration diagram of a remote monitoring maintenance system according to a third embodiment. Fig. 10 is an operation flowchart at a time when the remote monitoring maintenance system according to the third embodiment classifies device information. Fig. 11 is an operation flowchart at a time when the remote monitoring maintenance system according to the third embodiment selects a priority maintenance item. Fig. 12 is a hardware configuration diagram of a remote monitoring apparatus according to the first to third embodiments. Description of Embodiments [0012] First Embodiment Hereinafter, a remote monitoring maintenance system according to the first embodiment will be described with reference to Fig. 1. Fig. 1 is a configuration diagram of the remote monitoring maintenance system according to the first embodiment. [0013] In Fig. 1, the remote monitoring maintenance system includes a monitoring target device la, a monitoring target device lb, and a remote monitoring apparatus 2. The monitoring target devices la and lb are connected to the remote monitoring apparatus 2 via a network 100. The monitoring target devices la and lb are, for example, devices such as an elevator. The remote monitoring apparatus 2 is an apparatus that monitors the monitoring target devices la and lb, and calculates a failure risk of each part of the monitoring target devices la and lb. The failure risk will be described later. Note that, in the description below, there will be described an example in which the number of monitoring target devices to be monitored by the remote monitoring apparatus 2 is two monitoring target devices which are la and lb. However, the remote monitoring maintenance system can be realized even with two or more monitoring target devices. Hereinafter, the monitoring target devices la and lb and other monitoring target devices are collectively referred to as a monitoring target device 1. [0014] The remote monitoring apparatus 2 includes a device information storage unit 23, a maintenance information storage unit 24, a maintenance information input unit 25, a failure risk processing unit 26, and a maintenance item selection unit 27. [0015] The device information storage unit 23 stores a device information database (DB). The device information DB has device information transmitted from the monitoring target device 1. The device information is information transmitted from the monitoring target device 1 and corresponds to, for example, the number of times of activation, a travel distance, an accumulated operation time, and the like of the elevator. [0016] The maintenance information storage unit 24 stores a maintenance information DB. The maintenance information DB has maintenance information obtained by a maintenance worker upon visiting a site and inspecting a part included in the monitoring target device 1. The maintenance information is, for example, the remaining thickness of a brake shoe, an amount of gear oil or deterioration of a hoist, and the like of the elevator. [0017] The maintenance information input unit 25 is an input unit to input the details of a maintenance work performed by the maintenance worker at each site, as the maintenance information. [0018] Based on the details of the device information DB and the maintenance information DB, the failure risk processing unit 26 calculates the failure risk of the monitoring target device 1 at each site. The failure risk is an index for the maintenance worker to determine whether or not a maintenance item relating to any part among parts constituting the monitoring target device 1 is inspected, and is calculated for each part of the monitoring target device 1. [0019] The failure risk is calculated as the product of a failure probability obtained from the device information and the maintenance information and a risk value predetermined for each part. The failure probability obtained from the device information and the maintenance information is calculated by the failure risk processing unit 26 by calculating a failure probability based on the device information and correcting this failure probability based on the maintenance information obtained by the maintenance worker at a previous inspection. [0020] It is generally known that the failure probability increases as time elapses. The failure probability calculated based on the device information may be calculated using a general method such as Weibull distribution. With respect to a method of correcting the failure probability based on the maintenance information, for example, a threshold value relating to the maintenance information at the previous inspection may be set, and the failure probability may be corrected depending on whether a value indicated in the maintenance information is equal to or larger than the threshold value or equal to or smaller than the threshold value. Specifically, the failure risk processing unit 26 may perform in such a way that if the maintenance information at the previous inspection is equal to or smaller than the threshold value, the failure probability is reduced by 10%, and if equal to or larger than the threshold value, the failure probability is added by 10%. That is, the failure risk processing unit 26 corrects the failure probability at the next inspection calculated based only on the device information with a correction value in accordance with the threshold value, and sets the corrected failure probability as a failure probability at a next inspection. Therefore, when it is determined, from the maintenance information obtained by the maintenance worker at the previous inspection, that the degree of friction of a brake and a rope is low and margins of remaining portions are wide, the failure probability relating to the part is corrected to be low. Note that the failure risk processing unit 26 according to the present embodiment does not necessary need to calculate the failure probability using the threshold value, for example, a relational expression between the maintenance information and the correction value of the failure probability is predetermined, and the failure probability may be corrected based on the maintenance information. The above threshold value and the relational expression may be determined based on simulation and an empirical rule. [0021] The risk value is a value predetermined for each part and is set in accordance with a risk at a time when the part fails. The risk value is the value that can be arbitrarily changed by a designer or an administrator of the remote monitoring maintenance system, and can be set from various viewpoints such as the danger of an occupant, the magnitude of device damage, and a repair cost in a case where the part fails. As just described, since the failure risk processing unit 26 calculates the failure probability based on the maintenance information in addition to the device information, it is possible to calculate the failure probability more accurate than the failure probability calculated based only on the device information. [0022] A configuration of the failure risk processing unit 26 will be described in detail with reference to Fig. 2. Fig. 2 is a functional block diagram of the failure risk processing unit 26. In Fig. 2, the failure risk processing unit 26 includes a failure probability calculation unit 261, a failure probability correction unit 262, a risk value storage unit 263, and a failure risk calculation unit 264. [0023] The failure probability calculation unit 261 acquires the device information from the device information DB in the device information storage unit 23, and calculates the failure probability for each part. [0024] The failure probability correction unit 262 acquires the maintenance information from the maintenance information DB stored in the maintenance information storage unit 24, and based on this maintenance information, also corrects the failure probability calculated by the failure probability calculation unit 261. [0025] The risk value storage unit 263 stores the risk value for each part. [0026] The failure risk calculation unit 264 calculates the failure risk based on the failure probability for each part corrected by the failure probability correction unit 262 and the risk value stored in the risk value storage unit 263. [0027] In Fig. 1, based on the failure risk calculated by the failure risk processing unit 26, the maintenance item selection unit 27 selects a maintenance item (a priority maintenance item) relating to a part requiring an inspection at the next inspection. Upon selection of the priority maintenance item, the maintenance item selection unit 27 selects the maintenance item so as to preferentially maintain those having high failure risk. For example, the maintenance item selection unit 27 calculates failure risks which are a failure risk X at the next inspection and a failure risk Y at an inspection after the next inspection. Assuming that a threshold value of the failure risk requiring maintenance is A, there are the following three patterns. (1) X apparatus 2 and other device (the maintenance item selection apparatus 3 or the like) are separated as illustrated in Fig. 5. The storage device 203 is a device that executes functions of the device information storage units 23, 23a, 23b, and 23c, the maintenance information storage unit 24, the failure history storage unit 29, the working proficiency level storage unit 51, and the risk value storage unit 263. The processing device 204 is, ) for example, a Central Processing Unit (CPU) and executes functions of the failure risk processing unit 26, the maintenance item selection unit 27, the similar device classification unit 28, the working proficiency level storage unit 51, the maintenance work optimization unit 52, the failure probability calculation unit 261, the failure probability correction unit 262, and the failure risk calculation unit 264. Note that each i function of the device information storage unit 23, the failure risk processing unit 26, the maintenance item selection unit 27, the similar device classification unit 28, the maintenance work optimization unit 52, the failure probability calculation unit 261, the failure probability correction unit 262, and the failure risk calculation unit 264 may be configured by hardware, and may be realized by allowing a computer such as the ) processing device 204 to execute a predetermined program. Reference Signs List [0072] 1, la, lb: monitoring target device, 2: remote monitoring apparatus, 3: maintenance item selection apparatus, 4: external server, 23: device information storage unit, 24: maintenance information storage unit, 26: failure risk processing unit, 27: i maintenance item selection unit, 28: similar device classification unit, 29: failure history storage unit, 51: working proficiency level storage unit, 52: maintenance work optimization unit, 261: failure probability calculation unit, 262: failure probability correction unit, 263: risk value storage unit, 264: failure risk calculation unit, 100: network, 201: input device, 202: output device, 203: storage device, 204: processing device [Claim 1] A remote monitoring apparatus comprising: a failure probability calculation unit to calculate a failure probability for each part of a monitoring target device at a next inspection, based on device information acquired from the monitoring target device; a failure probability correction unit to correct the failure probability based on maintenance information at a previous inspection by a maintenance worker who inspects the monitoring target device; and a failure risk calculation unit to calculate a failure risk at the next inspection based on the failure probability corrected by the failure probability correction unit. [Claim 2] The remote monitoring apparatus according to Claim 1, comprising a maintenance item selection unit to select, among parts of the monitoring target device, a maintenance item relating to a part whose failure risk is calculated by the failure risk calculation unit exceeds a threshold value, as a priority maintenance item requiring an inspection work at the next inspection. [Claim 3] The remote monitoring apparatus according to Claim 2, wherein the failure risk calculation unit calculates a failure risk at an inspection after the next inspection for each part of the monitoring target device, based on the device information and the maintenance information, and wherein the maintenance item selection unit selects the priority maintenance item based on the failure risk at the next inspection, the failure risk at the inspection after the next inspection, and the threshold value. [Claim 4] The remote monitoring apparatus according to Claim 1, wherein the failure risk calculation unit calculates the failure risk based on the failure probability corrected by the failure probability correction unit and a risk value of said each part. [Claim 5] The remote monitoring apparatus according to Claim 2 comprising a working proficiency level storage unit to store proficiency of maintenance workers, wherein the maintenance item selection unit selects, based on the proficiency of maintenance workers stored in the working proficiency level storage unit, a maintenance worker who is assigned to the inspection work of the selected priority maintenance item. [Claim 6] The remote monitoring apparatus according to Claim 4 comprising a failure history storage unit to store failure history of each part, wherein the failure risk calculation unit sets the risk value based on the failure history stored in the failure history storage unit. [Claim 7] A remote monitoring maintenance system comprising: a device information storage unit to sore device information acquired from a monitoring target device; a maintenance information storage unit to store maintenance information at a previous inspection by a maintenance worker who inspects the monitoring target device; a failure probability calculation unit to calculate a failure probability for each part of the monitoring target device at a next inspection, based on the device information; a failure probability correction unit to correct the failure probability based on the maintenance information; a failure risk calculation unit to calculate a failure risk based on the failure probability corrected by the failure probability correction unit; and a maintenance item selection unit to select, among parts of the monitoring target device, a maintenance item of a part whose failure risk calculated by the failure risk calculation unit exceeds a threshold value, as a priority maintenance item requiring an inspection at the next inspection. [Claim 8] A remote monitoring method comprising; a failure probability calculation step to calculate a failure probability for each part at a next inspection, based on device information acquired from the monitoring target device; a failure probability correction step to correct the failure probability based on maintenance information at a previous inspection by a maintenance worker who inspects the monitoring target device; and a failure risk calculation step to calculate a failure risk based on the failure probability corrected by the failure probability correction step. [Claim 9] A remote monitoring program to cause a computer to execute: a failure probability calculation step to calculate a failure probability for each part of a monitoring target device at a next inspection, based on device information acquired from the monitoring target device; a failure probability correction step to correct the failure probability based on maintenance information at a previous inspection by a maintenance worker who inspects the monitoring target device; and a failure risk calculation step to calculate a failure risk at the next inspection based on the failure probability corrected by the failure probability correction step.